Social Media DisruptionNigeria's WhatsApp Politics
How are social media and digital technology shaping elections? This question is more important than ever, yet few studies look at WhatsApp's impact on the political landscape—even in Africa, where it is the dominant messaging platform. This article combines a case study of Nigeria's 2019 elections with surveys and analysis from Ghana, Kenya, Malawi, and Sierra Leone to show how social media are used by parties, candidates and voters. We conclude that WhatsApp is a disruptive technology that challenges existing hierarchies in ways that are simultaneously emancipatory and destructive, strengthening and undermining democratic consolidation at the same time.
How are social media and digital technology shaping elections? In today's political environment, this question is both more urgent and more controversial than ever before. As internet penetration and access to smartphones expand, a growing share of voters are receiving information about politics via Twitter, Facebook, and a raft of digital messaging platforms. In most African countries, WhatsApp—a Facebook-owned app that enables users to share text messages, images, and voice notes over the internet—has become the default mode of telecommunication, especially in urban areas. Because WhatsApp is encrypted, it is also favored by government operatives trying to spread messages surreptitiously, and by opposition activists who seek to evade government controls in competitive authoritarian states such as Uganda and Zimbabwe.1
The growing centrality of WhatsApp to both everyday life and political communication has led to a frenzy of media coverage. Much of this has focused on WhatsApp as a potential channel for what is usually referred to as "fake news"—though this term often obscures as much as it reveals. In the space of just a year, countries as otherwise diverse as Brazil, India, and Nigeria were said—with varying degrees of accuracy—to have witnessed their first "WhatsApp election," with the dissemination of rumors, conjecture, and lies allegedly undermining the democratic process itself.2
Current debates about WhatsApp reflect changing attitudes concerning the impact of new digital technologies on democracy more generally. In the past decade, the number of commentators skeptical about the democratic potential of these technologies has grown, and the nature of [End Page 145] their skepticism has also changed. Larry Diamond channeled the hopes of many when he wrote in 2010 of "liberation technology": any form of information and communication technology (ICT) that "can expand political, social, and economic freedom." Diamond went on to point out that there were two opposing schools of thought regarding the impact of computers, mobile phones, and the internet in entrenched authoritarian states. For optimists, "liberation technology enables citizens to report news, expose wrongdoing, express opinions, mobilize protest, monitor elections, scrutinize government, deepen participation, and expand the horizons of freedom." By contrast, pessimists viewed liberation technology as relatively powerless in the face of authoritarian censorship and repression. This debate reflected an assumption that the key question was who would end up controlling technology, and whether "people, organizations, and governments" could effectively mobilize to ensure freedom of access to digital content.3 It was generally assumed that if they succeeded in this, technology would strengthen democracy.
Over the last decade, the focus of the debate has evolved dramatically. Today, there are many more pessimists, and they tend to be more concerned about the risk that unfettered access to ICT will undermine or has already undermined democracy than about the risk that authoritarian regimes will be able to turn it off. Rather than exploring the prospects for digital liberation, many researchers and policy makers have shifted to focusing on how "the success of social media has fueled political turmoil in democracies."4 It is commonplace to hear, for instance, that Twitter and Facebook are lending themselves to the creation of "echo chambers" and the spread of misinformation. Such concerns are even more acute when it comes to WhatsApp, whose encrypted nature some believe encourages the spread of fabricated and exaggerated stories that can stoke ethnic and political tensions. As a result, the platform stands accused of undermining public confidence in political systems while raising the risks of conflict and instability. After having initially worried about the impact of government censorship, some democrats have begun suggesting that there needs to be greater control over the flow of information online.5 In short, the question is now whether digital technology is itself the problem.
Understanding WhatsApp's Impact
Yet for all the column inches that have been devoted to this topic, there have been few studies of how WhatsApp actually works and its impact on the political landscape, even in Africa where it is the dominant messaging platform in forty countries. This is in part because WhatsApp is very hard to research—its encrypted nature means that one cannot easily track how messages are shared, or who shares them. Thus we are often left guessing about how Africa's most important digital messaging [End Page 146] platform is affecting democracy. To start to build a more accurate picture, we researched the role of WhatsApp in the Nigerian general elections of 2019. Nigeria makes for a particularly interesting case study: One of Africa's largest and most populous nations, it is ethnically and religiously diverse and exhibits many features of a classic competitive authoritarian state. Freedom House's 2019 Freedom on the Net report assigns Nigeria a rating of Partly Free, and there are complaints about the excessive use of force and censorship by state actors against both political opponents and social-media critics.
In order to understand both how political parties seek to use WhatsApp and how messages are received by Nigerian voters, we balanced 46 interviews with party activists and social-media "influencers" against seven focus groups composed of a range of different demographics and a unique survey of a thousand urban residents in two states. To make sure that we did not misinterpret trends in a single region as nationally representative, we held interviews in three states: Abuja, the capital; Kano, in the North West; and Oyo in the South West. While these locations are not fully representative of Nigeria's diversity, they did allow us to leverage the variation between Kano, where ethnic and religious tensions are particularly pronounced, and Oyo, where they have historically been more contained. Finally, to encourage participants to speak openly about activities that could in some cases be viewed as unethical or criminal, we offered them full anonymity.
Viewed together with surveys and analyses from Ghana, Kenya, Malawi, and Sierra Leone, our research demonstrates that it is a mistake to see WhatsApp as simply generating "political turmoil" in Africa, or to set up "turmoil" and "liberation technology" as competing explanations of how online communication and debate are changing politics. It is true that in Nigeria, and elsewhere in Africa, WhatsApp has emboldened candidates and their supporters to spread "fake news"—whether inadvertently (misinformation) or deliberately (disinformation). In some cases, this has eroded trust in established parties and exacerbated intergroup tensions, increasing the risk of political violence. Yet at the same time, WhatsApp has created new opportunities for women and young people to engage in politics, enabled better coordination by opposition parties and civil society groups, and undermined government efforts to control the flow of information. In other words, WhatsApp is a disruptive technology that challenges existing hierarchies in ways that are simultaneously emancipatory and destructive, strengthening and undermining democratic consolidation at the same time. The challenge is therefore to understand both aspects of WhatsApp's impact and the ways in which they interact.
Our research points to several key findings. First, politicians and parties must harness existing social and political structures or create new ones in order to effectively direct messaging on WhatsApp and social [End Page 147] media, as Nigeria's political leaders are trying to do. Second, while digital technologies are certainly being used to spread false information, "fake news" often proves ineffective; misleading stories are far likelier to gain traction if they are built on an element of truth. Third, the growing importance of social media has enabled a younger cohort of influencers to take on a more prominent political role, while also creating new space for women to voice their opinions and empowering opposition parties to outmaneuver the security forces and contest government propaganda.
Overall, WhatsApp has had an uneven impact on democracy in Africa, presenting new opportunities for some historically marginalized groups while also raising the risk of political instability. Although there is growing pressure for governments to step up regulation in response to the challenges generated by new technology, such demands are illadvised in electoral authoritarian settings such as Nigeria. Enabling governments with dubious democratic credentials to more effectively censor WhatsApp would not stop the flow of "fake news"—it would simply grant the ruling party monopoly control over its spread.
The Use and Abuse of WhatsApp
Nigeria's 2019 presidential election pitted serving president Muhammadu Buhari of the All Progressives Congress (APC) against Atiku of the People's Democratic Party (PDP). Both in their seventies, the two men were no strangers to Nigerian voters. Buhari had seized power as a military dictator in the 1980s and had run—as a "converted democrat"—in every Nigerian election since 2003.6 Abubakar, universally referred to as "Atiku" in Nigeria, served as the country's vice-president between 1999 and 2007, having won the office on a PDP ticket. Over the following decade, he sought the presidential nomination from both the PDP and the APC (founded in 2013). With some local variation, APC-PDP competition was the key electoral dynamic at the state as well as the federal level in 2019.7 Buhari won the presidential election, held in late February, with support from 56 percent of voters.
Although 2019 was not, in fact, Nigeria's "first" WhatsApp election, the app played a much more central role in political parties' ground game than it had four years previously.8 In an interview, an Ibadan-based APC social-media advisor active in both 2015 and 2019 explained that previously parties and their agents had used the platform in a fairly ad hoc fashion, and Facebook was the preferred social-media app for campaigning. Between the two polls, however, a consensus emerged across the political establishment that WhatsApp had become a critical means of mass communication, political organization, and electioneering. Discussions held in April 2019 among our focus groups in Kano and Ibadan (the capital of Oyo) suggest that this conclusion was informed partly by the platform's [End Page 148] rapidly growing popularity and partly by its accessibility, affordability, (relative) anonymity for users, and low barriers to entry.
The investments that Nigerian political leaders made in harnessing WhatsApp underscore the level of effort required to successfully exploit this platform. While the speedy circulation of messages often tempts commentators to talk of WhatsApp "memes" as a spontaneous phenomenon, this is misleading. It is of course true that messages can take on a life of their own, but to exert influence over WhatsApp traffic requires an effective structure. Unlike posts on Twitter, WhatsApp messages can only be sent to people who are among your contacts, and, while it is possible to form WhatsApp groups, these are capped at 256 people. As a result, effectively communicating through this medium requires setting up multiple overlapping groups, a task that demands significant organization. In some cases, Nigerian parties and their supporters can coopt existing WhatsApp networks built around church or student-union groups, but political players have also felt the need to invest in carefully designed infrastructure so as to be sure that their messages are getting across. Thus WhatsApp does not simply take the place of party and social organizations. Instead, effective use of the platform actually depends on the existence of these structures, and this gives ruling parties—which typically enjoy greater access to campaign funding and state resources—a built-in advantage.
By 2017, for example, an APC-aligned Buhari New Media Centre (BNMC) had been established in Abuja. The BNMC focused on overseeing, and in some cases establishing, networks of locally embedded WhatsApp groups across all 36 of Nigeria's states. Some of these groups had been created in the lead-up to the 2015 election to promote APC candidates. In other cases, APC mobilizers "grew" their own clusters of groups. For each state, the BNMC National Committee selected a "chapter" leader who, following some training, would update the center about developments on the ground and share messages—some crafted in Abuja—among his or her own local networks. Such networks, woven together by activists who built on existing WhatsApp groups set up by school alumni, churchgoers, and so forth, have an extensive social and geographic reach that makes them a powerful asset for campaigns. As one BNMC official noted, "in less than ten minutes, information can spread across the country."9
The opposition PDP set up a similar structure, the Atikulated Youth Force (AYF), but with fewer resources at its disposal this entity struggled to achieve the same degree of nationwide coverage as the BNMC. This limitation is even more pronounced for smaller parties that struggle financially, a pattern also observed elsewhere in Nigeria's neighborhood—for instance in Ghana, where "social media is not currently closing the gap between the 'big two' [political parties] and the smaller parties and might even be making it worse."10 [End Page 149]
As our interviewees related, the AYF's members reported to an eighteen-member Presidential Committee on Social Media that, like the BNMC, shaped campaign messages, fed them down the chain, often through WhatsApp groups, and fashioned responses to attacks and negative stories. Both the BNMC and AYF remained technically separate from the parties they supported, though in practice they were closely linked in terms of coordination and personnel. "For the most part, we are left to our own devices," noted one BNMC Kano representative in our interviews, "but when there is a critical issue we do occasionally receive a directive from above that we are encouraged to follow." Likewise, one Abuja-based AYF official explained that "I have access to the official campaign team but we are separate from it . . . sometimes there is overlap and we work together but there is not a joint approach."
Even with this considerable investment in social-media coordination, the ability of national, party-aligned structures to "control the message" at the local level proved limited. During the 2019 polls, these difficulties were compounded as parties with limited resources found themselves in tight competition for the services of tech-savvy entrepreneurs and social-media influencers. In many cases, administrators of state-level or local WhatsApp groups could expect only limited funds (to purchase data)—as well as the prospect of future employment if their preferred candidate was victorious.
Coordination outside the presidential campaigns—for example, at the gubernatorial level—was even more improvised, contingent on available resources, and driven by individual candidates rather than parties. The cooptation of existing social networks—embedded in rhizomes of overlapping WhatsApp groups—was a central strategy of party activists, with reams of personal contacts often added to party-aligned groups without individuals' consent. Negotiating the support of social-media influencers—from media personalities to actors and musicians—was also a common approach. "We push messages to influencers, who have wide networks," noted one gubernatorial-campaign aide, with another explaining that their campaign contacted Nollywood actresses and other celebrities; "some may have negotiated a role in the government as advisors."
The degree of coordination between parties and their WhatsApp promoters—self-styled "propaganda secretaries," as members of one of our Kano focus groups put it—and overall sophistication of campaigns' social-media operations varied not only across levels but also between states. This was a matter both of resources and of self-protection: By retaining official distance between themselves and the social-media activists who operate on their behalf, candidates can maintain plausible deniability regarding any messages that prove particularly controversial. As our focus-group participants in Kano also noted, "[political parties and candidates] tried to keep their distance . . . they allow social-media supporters to do their dirty work." [End Page 150]
It is therefore important to be realistic about the capacity of political parties to use WhatsApp for political mobilization—especially as Nigeria appears to be a regional leader in this regard. A study of the use of WhatsApp in Sierra Leone's 2018 election noted that only one political party employed staff (three people) to manage its social-media operations.11 Under these conditions, messaging on WhatsApp is likely to supplement rather than replace existing campaign strategies.
Fake News and Its Limits
While ruling parties may have the easiest time establishing effective structures, this does not mean that such parties are always best placed to control the flow of information on WhatsApp. How widely stories circulate also depends to a great degree on whether they resonate with the wider population. Although most campaign aides and consultants insisted that they only used WhatsApp to share reliable information about their candidate's policies, it is clear that they also produced messages critiquing their opponents—what in the United States would be called "attack ads." In some cases, this entailed spreading "fake news." One local councillor and social-media advisor to a gubernatorial campaign in Oyo acknowledged that he and counterparts elsewhere would "cook one or two stories a day" and then "put it on WhatsApp . . . especially when it has to do with character assassination." More often, though, political parties claimed to be above such tactics while accusing opponents of manufacturing "fake news"—about one another, the government, or the election commission. In January 2019, for example, the Buhari campaign accused Atiku of "sponsor[ing] fake news aimed at attacking the morale and credibility of Nigeria's armed forces."12
As the effective dissemination of "fake news" via WhatsApp to benefit successful opposition campaigns in Brazil (2018) and the United States (2016) has demonstrated, however, this is one area in which the ruling party may operate at a disadvantage. Precisely because voters frequently view the government's official news output with great cynicism, they are often willing to believe rumors—even outlandish ones—about the ruling party and the president. The most widely shared story in Nigeria in the run up to the 2019 election alleged that President Buhari had died and been replaced by a clone—a lookalike from Sudan named Jubril.13 So pervasive was this rumor that Buhari felt compelled to publicly deny it; "it got to the point," noted one presidential social-media aide whom we interviewed, "where the only person who could dispel it was the president himself. It started off as an organic campaign but became increasingly sophisticated."
The spread of the Buhari clone story reflects widespread popular cynicism about the veracity of government news following a series of scandals and cover-ups. The meme also played effectively on the particularities [End Page 151] of Nigeria's recent history. A decade before the 2019 polls, one of Buhari's predecessors—Umaru Yar'Adua—had been flown abroad for medical treatment, just as Buhari was during his first term. In both cases, the state of the president's health was kept largely secret from the general population. Whereas Yar'Adua died ten weeks after his return to Abuja in February 2010, however, Buhari came back after several months abroad seemingly healthier than ever. Yar'Adua's fate, a lack of transparency about the president's health, and the rarity of seeing an elderly politician make a full recovery all combined to make the clone rumor more credible.
The spread of the clone story demonstrates that "fake news" is most likely to be believed when it contains an element of truth or reflects assumptions and prejudices already held by the electorate. Messages that fail to do this may be widely ignored, especially given that voters are well aware of WhatsApp's limitations as a source of reliable information. A survey conducted in Kenya in 2017 found that only 18 percent of people trusted WhatsApp "a lot," as opposed to 45 percent for radio and 49 percent for television.14 Our interviews and focus groups revealed similar caution in Nigeria.
This nuance helps to explain why some stories spread much more extensively than others. In the aftermath of the gubernatorial races in Kano and Oyo, for example, our survey asked respondents whether they had seen rumors on social media that the unsuccessful candidate had, in fact, won the election. In Kano, 91 percent of respondents had heard this story, with 75 percent encountering it via social media (84 percent of these first on WhatsApp). In Oyo, however, only 8 percent had come across the rumor. Clearly, a host of variables—from digital literacy to local campaign dynamics—need to be considered when making sense of such discrepancies. But what the most pervasive stories have in common is that they resonate with existing understandings of how individuals and the political system more generally operate. Focus-group respondents in Kano noted, for example, that people were primed to believe rumors of the deputy police commissioner taking bribes after a video was released that showed the state governor receiving bribes from a contractor.
A secondary—but nonetheless significant—factor in the success of messages is whether authority figures lend them credibility. In December 2018, for example, prominent preacher David Oyedepo commended the Buhari clone story to his vast congregation.15 Indeed, our focusgroup [End Page 152] discussions in both Kano and Ibadan suggest that religious leaders have played a critical role in amplifying misinformation and disinformation, their voice carrying as much weight online as it does in physical community life. One focus group participant in Ibadan recalled that his pastor had posted an erroneous story in a church WhatsApp group regarding Atiku's purported victory in the presidential election. When the participant in question corrected the error, he was contacted privately by the pastor and scolded: "You are the only one who always challenges me [as to whether] something is true."
The intimate and "echo-chamber" character of nonpolitical WhatsApp groups may also discourage internal debate over the accuracy of particular stories. "Unless someone else that [a person] trusts more comes with a different truth," noted one scholar whose work has focused on social-media engagement in Nigeria, "they won't stop believing the original content."16 This is especially the case for groups whose members are already associated, politically, socially, and personally. "There is a personal nature to connections on WhatsApp," explained one local politician and campaign social-media advisor in an interview, "that is why WhatsApp is more useful."
Critically, then, WhatsApp forms part of a wider ecosystem of "fake news," building on a well-established offline "rumor mill" that first grew out of censorship during Nigeria's decades of military rule.17 Some stories originated outside the platform but "went viral" when transferred online, while others began life on social media but spread through a range of other mechanisms, including personal interactions, radio phone-in programs, and sermons. In other words, WhatsApp was one link in a chain of information-sharing mechanisms that criss-crossed the digital and physical worlds; it amplified, but did not fundamentally alter, the creation and sharing of "fake news."
There are efforts to counter the spread of false information, including the good work of organizations such as Fact Check Nigeria. Many WhatsApp administrators also have taken a strict approach to the sharing of political news in groups established for other purposes. A medialiteracy activist in Abuja who administered a group for the whistleblowing organization Corruption Anonymous explained that, "those who join the group are informed of the ground rules at the point of invitation—no jokes, no religious content, no political content—though people don't always listen to that."18 Both these strategies have a limited effect, however. Fact-checking organizations may not see all of the unreliable stories circulating on WhatsApp and lack the capacity to reply to every one. Even when they do respond, only a very small proportion of those who see the original story are likely to see the correction.
For their part, administrators are challenged by the large size of WhatsApp groups, which makes rules difficult to enforce. In our survey of Nigeria, 73 percent of respondents reported that the average [End Page 153] size of groups they belonged to exceeded fifty members. According to the social-media director of one of the campaigns, some aides found even the 256 group-member limit imposed by WhatsApp to be too low, leading them to turn to other platforms such as Telegram. In many cases, therefore, the sheer volume of information being shared in and across WhatsApp groups makes it difficult to effectively police content, even in extremely problematic cases. One example of this, mentioned by our focus-group participants and a social-media advisor in Ibadan, was evident on Nigeria's election day: Pictures of political violence—some real, some taken at other times or in other places—spread rapidly across WhatsApp groups in both Kano and Ibadan, heightening tensions and lowering voter turnout.
A Disruptive Force
WhatsApp is best thought of as a disruptive technology that has the potential to be both emancipatory and destructive. For ruling parties, this medium presents some advantages—such parties generally have an easier time setting up the kind of structures needed to take full advantage of the platform—but also important challenges, insofar as they often struggle to control the spread of "fake news" that is detrimental to their interests. In addition, WhatsApp is challenging existing hierarchies by creating new opportunities for younger figures to participate in campaigns, giving a greater voice to women, and creating a "safe space" for opposition and civil society groups to plan and organize.
The growing importance of WhatsApp in African elections has helped to amplify the voices of more marginalized constituencies. Because securing political office in Nigeria depends on access to deep pockets and to powerful political patrons or "godfathers," there has been limited scope for involvement by younger people without private wealth at their disposal.19 The growing importance of WhatsApp, however, has opened up new space for younger, tech-savvy Nigerians to influence campaign strategy. As one presidential social-media advisor observed, "In campaigns at the higher levels there is very little digital savvy so it leaves a lot of space for younger people who are using it for leverage . . . social media is where young people gain an edge in politics."
Thanks to these dynamics, a cadre of young, freelance politico-digital entrepreneurs have inserted themselves into national- and state-level campaigns by trading on their social-media expertise and offering up WhatsApp group networks that they have assembled. Few of our respondents in this category suggested that they had been paid directly by political parties for services along these lines, though many explained that they had received "in-kind" support ranging from (most often) airtime to phones or tablets. In some cases, even more imaginative remuneration systems were devised. Interviewees from the AYF noted that seven-thousand members, for example, [End Page 154] were rewarded for being particularly prolific messengers by being named "AYF Ambassador of the Week," receiving a campaign T-shirt and in some cases the chance to meet Abubakar himself.
More often, engagement in campaigns' social-media outreach was motivated by hopes of future reward. "Many people hope or expect," noted one Kano-based PDP social-media aide, "that when—or if—the government is elected that they would either be offered a job in the administration, scholarships to study abroad, or small contracts to do work for the state government." While this means that rich and powerful patrons remain crucial, since social-media entrepreneurs look to them for professional advancement, the growing significance of digital technology does appear to be enabling young strategists to take on more prominent roles. This trend is likely to continue in Nigeria and beyond, given the much greater engagement of younger people with social media: A 2017 survey in Kenya, for example, found that 41 percent of 18 to 24 year olds in the country have WhatsApp accounts, with this figure declining to less than 10 percent for those older than 45.20
WhatsApp has also enabled women to increase their participation in political discussion, especially in more conservative communities. Participants in an all-female focus group in Kano explained that WhatsApp groups have provided a range of opportunities for women—from learning about and communicating with politicians to forming, through particularly large and influential groups, an effective "vetting committee" for local candidates. "Women rely more on WhatsApp than men," noted one respondent, " . . . in conservative households they don't have the opportunity to move around and see things with their own eyes, but their phone—with pictures and videos—can allow them to do that." These sentiments align with conclusions drawn by other researchers, who have found that WhatsApp allows women to "to push the barriers of societal norms that typecast them in gendered roles as mothers and housewives . . . to become agents of their own change."21
WhatsApp has also strengthened the hand of opposition parties and civil society groups that have historically faced repression and harassment. Opposition candidates, while recognizing that WhatsApp carries stories that threaten their reputation, were quick to point out how this medium shifted the balance of political communication in their favor. One opposition contender for a local-council seat in Ibadan compared the situation with WhatsApp favorably to a former era in which the government monopolized the airwaves. In particular, he explained that whereas there used to be no "right of reply" to ruling-party propaganda spread by state-controlled radio and newspapers, WhatsApp made it possible to counter new fake stories within seconds—and, of course, to disseminate antigovernment messages. This was not an ideal situation, he admitted, but it did "democratize" the field of political communication.
Opposition candidates and civil society groups also emphasized the [End Page 155] usefulness of WhatsApp as a tool of political coordination. Especially at the subnational level, candidates valued the platform more for its encrypted nature, which meant they could safely use it to discuss campaign strategy with their advisors, than for its potential use in mobilizing voters. While political actors are wary of the risk of "moles" infiltrating their groups, they see WhatsApp as significantly safer than sending regular texts, emails, or voice calls. This echoes the experience of human-rights organizations and opposition parties in more repressive environments such as Uganda and Zimbabwe, which used WhatsApp to evade state surveillance while assembling evidence of government misdeeds around recent elections. Although governments retain the power to turn off internet access, and hence WhatsApp, they are generally reluctant to rely on this tactic over long periods due to the heavy economic toll it takes.22 As a result, the platform has strengthened the position of watchdog bodies and opposition parties.
Between Liberation Technology and Political Turmoil
It is clear from the Nigerian elections of 2019 that WhatsApp is operating as "liberation technology" and contributing to "political turmoil" at the same time. These are not mutually exclusive outcomes, but rather two sides of the same coin. WhatsApp's disruptive capability, highly valued by opposition parties and civil society groups, also facilitates the spread of "fake news." This means that in any given election WhatsApp is likely both to strengthen democracy, by equalizing access to information and facilitating the coordination of opposition and civil society efforts, and to challenge it, by undermining public trust in politicians and the wider political process. Whether it is the platform's emancipatory aspect or its destructive potential that comes to the fore, however, will vary across political contexts. The emancipatory effect of WhatsApp will be most significant in authoritarian countries marked by repression and censorship. The destructive potential of fake news, on the other hand, is likely to be greatest where digital literacy and public trust in political institutions are low, and where intercommunal relations are strained.
"Fake news" spread by WhatsApp had a notably different impact in Kano, which has historically suffered from riots and clashes that have at times followed a religious fault line, than in Oyo, which is a more ethnically homogenous state and has less of a history of interethnic and religious conflict. While there was limited evidence that fake news had inspired unrest in Oyo, the potentially destabilizing impact of WhatsApp in places like Kano is a cause for concern—especially as messages spread through the platform often reach far beyond the immediate circle of users. Our urban respondents noted that particular WhatsApp users could become influential sources of misinformation for communities and networks where internet penetration was more [End Page 156] limited. In Ibadan, focus-group respondents highlighted a local shopkeeper who "is the main purveyor of 'fake news,' she spreads it around to customers." Something similar has been observed in Malawi, where, although only 14 percent of the population has direct access to the internet, those linked in to social media may "pick up fake information and end up discussing these emerging issues in their neighborhoods, social gatherings, and in many public spaces."23
The challenge for those advocating reform of WhatsApp is that the obvious strategies to reduce its destructive potential would also undermine its emancipatory power. Removing the encryption from WhatsApp or expanding the powers of governments to track messages (and thus to prosecute those circulating fake stories) would also undermine the privacy that is so valued by opposition parties and civil society groups. None of the opposition leaders to whom we spoke favored this course of action, which they thought would not eradicate "fake news" but would simply give the government a monopoly on purveying information. Similarly, forcing individuals to use their real names in order to allow the tracking of messages would undermine the anonymity that enables female users to speak their minds more freely. Such a move might also make it harder for individuals to quietly exit groups to which they do not wish to belong.
Many of our respondents did, however, suggest modifications to the platform that, if adopted by WhatsApp management, could make it easier for people to avoid or disengage from groups created, or hijacked, for nefarious purposes. In particular, respondents favored making it harder for administrators to add members to groups without their permission, and possible for people to leave groups without alerting the rest of the members. "Often you get added to groups without consent," noted one focus-group respondent in Ibadan, "and then in religious groups you may get denigrated if you leave, with people saying that you are not a Christian."
Several platform changes adopted since the completion of our fieldwork demonstrate that adaptations to address the more problematic consequences of WhatsApp use are indeed possible. WhatsApp has introduced a feature allowing users to change their settings so that others cannot add them to groups, or so that only certain people (for instance, those on the user's list of contacts) can do so. It remains the case, though, that when a group member leaves, a message visible to all members appears in the group alerting them. More recently, the spread of false information concerning the coronavirus pandemic prompted WhatsApp to restrict the sharing of messages that have been forwarded repeatedly (five times): Whereas such messages could previously be passed on to five groups simultaneously, now users can share them with only one group at a time.24
Irresponsible statements by government leaders complicate the struggle against misinformation, however, sapping confidence in official information and blurring the dividing line between real and fabricated information. In this context, the most effective way to mitigate [End Page 157] the negative impact of "fake news" without neutering WhatsApp's emancipatory potential is through digital-literacy campaigns that encourage individuals to critically evaluate the messages they receive and to think harder about what they "forward." Not only will this reduce the harms flowing from "fake news" introduced on social media, it will also help to mitigate the power of "offline" rumors—and give citizens the tools they need to hold accountable any governments that fail to tell the truth. [End Page 158]
Nic Cheeseman is professor of democracy at the University of Birmingham.
Jonathan Fisher is reader in African politics and director of the International Development Department at the University of Birmingham.
Idayat Hassan is director of the Centre for Democracy and Development in Abuja.
Jamie Hitchen is an independent researcher.
This study was funded under WhatsApp's Research Awards for Social Science and Misinformation. WhatsApp have played no role in directing or managing the project—including its framing, inception, development, and implementation. Nor have they had any influence over the authors' analysis and delineation of findings. Previous versions of this article were presented at the Shehu Musa Yar'Adua Centre in Abuja (July 2019), the International Studies Association Conference at the University of Ghana in Accra (August 2019), and the Twenty Years of Democracy in Nigeria Conference at the University of Oxford (December 2019). The authors are very grateful to participants in these events for helpful feedback. The authors would also like to thank Mukhtar Bello, Hadiza Hassan, Emeka Njoku, and Joshua Akintayo for valuable research assistance provided in Ibadan and Kano from February through April 2019.
1. Admire Mare and Trust Matsilele, "Hybrid Media System and the July 2018 Elections in 'Post-Mugabe' Zimbabwe," and Hannah Muzee and Andrew Osehi Enaifoghe, "Social Media and Elections in Uganda: The Case of Bobi Wine and the Arua Primary Elections," both in Martin N. Ndlela and Winston Mano, eds., Social Media and Elections in Africa, Volume 1: Theoretical Perspectives and Election Campaigns (Cham: Palgrave Macmillan, 2020), 147–76; 195–213.
2. Billy Perrigo, "How Volunteers for India's Ruling Party Are Using WhatsApp to Fuel Fake News Ahead of Elections," Time, 25 January 2019, https://time.com/5512032/whatsapp-india-election-2019; "Brazil's WhatsApp Election Campaign," Christian Science Monitor, 27 September 2018, www.csmonitor.com/Commentary/the-monitors-view/2018/0927/Brazil-s-WhatsApp-election-campaign; "Nigerian Political Parties are Weaponising Fake News," Mail and Guardian (Johannesburg), 21 February 2019 https://mg.co.za/article/2019-02-21-nigerian-political-parties-are-weaponising-fake-news.
3. Larry Diamond, "Liberation Technology," Journal of Democracy 21 (July 2010): 70, 82.
4. Joshua A. Tucker, Yannis Theocharis, Margaret E. Roberts, and Pablo Barberá, "From Liberation to Turmoil: Social Media and Democracy," Journal of Democracy 28 (October 2017): 54.
5. Danielle Keats Citron, "What To Do About the Emerging Threat of Censorship Creep on the Internet," Cato Institute, 28 November 2017.
6. Muhammadu Buhari, in "Prospects for Democratic Consolidation in Africa: Nigeria's Transition," meeting transcript, Chatham House, London, 26 February 2015, www.chathamhouse.org/sites/default/files/field/field_document/20150226BuhariSpeechQA.pdf, 6.
7. Ayo Obe, "Aspirations and Realities in Africa: Nigeria's Emerging Two-Party System?" Journal of Democracy 30 (July 2019): 109–23.
8. Mala Mustapha, "The 2015 General Elections in Nigeria: New Media, Party Politics and the Political Economy of Voting," Review of African Political Economy 44, no. 152 (2017): 312–21.
9. The interviews cited in this manuscript were conducted in Abuja in February 2019, Kano in April 2019, and Ibadan in April 2019.
10. See Elena Gadjanova, Gabrielle Lynch, Jason Reifler, and Ghadafi Saibu, Social Media, Cyber Battalions, and Political Mobilisation in Ghana (Exeter: University of Exeter, 2019), 13.
11. Maggie Dwyer, Jamie Hitchen, and Thomas Molony, "Between Excitement and Scepticism: The Role of WhatsApp in Sierra Leone's 2018 Elections," in Maggie Dwyer and Thomas Molony, eds., Social Media and Politics in Africa: Democracy, Censorship and Security (London: Zed Books, 2019).
12. Dennis Erezi, "Group Accuses Atiku of Sponsoring Fake News Against Buhari," Guardian (Lagos), 3 January 2019.
13. "'It's the Real Me': Nigerian President Denies Dying and Being Replaced by Clone," Guardian (London), 2 December 2018.
14. Kenya Election Survey, Conducted by IPSOS-Kenya (2017) under the supervision of Nic Cheeseman, data available from the author.
15. Emmanuel Akinwotu, "Whose Truth? Disinformation and Misinformation Online in Nigeria," West Africa Insight 7, no. 1 (2019): 19.
16. Interview, Naima Hafiz Abubakar, Lecturer, Bayero University, by telephone, February 2019.
18. Interview, Chido Onumah, Coordinator, African Centre for Media Information and Literacy (AFRICMIL), Abuja, February 2019.
19. Mayeni Jones, "Nigeria Election 2019: How "Godfathers" Influence Politics," BBC, 4 February 2019, www.bbc.co.uk/news/world-africa-47089372; Ngozi Nwogwugwu, "Youth and Big Men Politics," in Kenneth Kalu, Olajumoke Yacob-Haliso and Toyin Falola, eds., Africa's Big Men: Predatory State-Society Relations in Africa (New York: Routledge, 2018).
20. Kenya Election Survey (2017).
21. Salihu Ibrahim Dasuki and Naima Hafiz Abubakar, "How Nigerian Women Are Using WhatsApp to Chat, Learn, and Earn," The Conversation, 30 October 2018, https://theconversation.com/how-nigerian-women-are-using-whatsapp-to-chat-learn-and-earn-105145.
22. Yomi Kazeem, "Internet and Social Media Shutdowns Cost African Economies over $2 Billion in 2019", Quartz Africa, 16 January 2020, https://qz.com/africa/1785609/internet-shutdowns-in-africa-cost-2-billion-in-2019.
23. Rabson Kondowe, "Despite a Low Internet Penetration, Malawi Is Worried About Fake News in its Election Run-Up," Quartz Africa, 10 May 2019 https://qz.com/africa/1616511/malawi-election-has-a-fake-news-problem-on-whatsapp-facebook.
24. Hadas Gold, "WhatsApp Tightens Limits on Message Forwarding to Counter Coronavirus Misinformation," CNN Business, 7 April 2020, www.cnn.com/2020/04/07/tech/whatsapp-misinformation-forward-limit/index.html.