Liberia’s Election 2023: Unmasking disinformation actors and channels

Getting your Trinity Audio player ready...

ABSTRACT

The study is a content analysis of fact-checks from three fact-checking organisations during and after the 2023 presidential elections in Liberia. The study found that Facebook remains the most popular social media platform for disseminating political disinformation. Even though the candidates of the two major political parties, George Weah and Joseph Boakai, were targets of political disinformation, the National Election Commission (NEC) was the biggest target of disinformation campaigns. The format for most of the mis/disinformation reports was largely verbal and pictorial text, with little attempt at video and audio manipulation.

1.0 INTRODUCTION

In a post-truth era characterised by the rapid dissemination of information through digital platforms such as social media, the essence of our democratic processes is being tested like never before. There are a plethora of examples of how disinformation campaigns are fast eroding the very foundations of democratic culture across Africa. According to Fortune Agbele in his report on “Disinformation and misinformation during Kenya’s 2022 election: Implications for voter confidence in the electoral process,” the twin impact of mis and disinformation constituted a “major dynamic during the August 2022 elections in Kenya.” At the heart of such campaigns is the role of media, particularly social media, with its increasingly blurred, overlapping influence between online and offline spaces in ways that are “not easy to track but important to recognise.” The Liberia 2023 general election provided yet another opportunity to ascertain the extent to which misinformation and disinformation campaigns were executed in pursuit of swaying public opinion and altering voting patterns. It is also a fact that fact-checking has been recognised by many credible institutions worldwide, including UNESCO, as an essential vehicle in fighting mis and disinformation. 

Against this backdrop, DUBAWA embarked on a mission to unmask the spreaders of misinformation and disinformation during the 2023 election season in Liberia and how fact-checking was used to mitigate such campaign impacts. The researcher also explored the crucial role of social media as a vehicle for this infodemic during the election.

Before we delve into this critical analysis, we must acknowledge the groundbreaking work DUBAWA had previously carried out in Liberia  before the country’s 2023 general elections. The report reinforced the notion about the prevalence of misinformation as a prominent feature in every democratic process. According to DUBAWA’s findings, supporters of George Weah, the presidential candidate of the Coalition for Democratic Change (CDC) appeared to be the principal agent spreading misinformation leading up to the election. That was before the October 10 general elections in Liberia. The findings in that report were thought-provoking and motivated the researcher to investigate what the actors would do during and after the election proper.

With that background, the researcher focused on election disinformation and explored how disinformation strategies were used in both the October 10 general elections and the November 14 run-off election, which saw the opposition Unity Party return to power.

2.0 SUMMARY OF LIBERIA’S ELECTION RESULTS  

It was a keenly contested election that had to take two rounds to determine a winner. In the first round, Weah secured 43.83% of the votes. At the same time, Joseph Boakai, his primary challenger from the Unity Party (UP), trailed closely with 43.44%, setting the stage for a runoff election scheduled for November 14, 2023. In the run-off, which suffered a low turnout, Boakai emerged as the winner with 50.89% against Weah’s 49.11%. The question to ask then is, were there examples of election disinformation during the two elections? Through what media, traditional or social, did the agents use in disseminating their disinformation campaigns? Which agents were associated more with mis and disinformation, and  what format did these disinformation campaigns take? 

The search and the desire to get answers to all these questions led DUBAWA to embark on a second explorative piece on election disinformation typology. Moreover, in a world where information is both a powerful tool and a dangerous weapon, it was imperative to unravel the intricacies of this critical chapter in Liberia’s democratic history and draw insights that can benefit not only African media enthusiasts but also citizens, policymakers, and stakeholders across the globe.

3.0 REPORT OBJECTIVES

This report aims to:

1. Examine the common media used in spreading misinformation at the Liberia 2023 election.
2. Identify prominent misinformation purveyors during and after the election. 
3. Examine the typology of misinformation during and after the Liberia 2023 election.
4. Identify which formats: texts, audio, or video were used more in the disinformation campaigns during elections.

4.0 METHODOLOGY

As a sequel to our previous publication, this study also analysed the contents of viral information that was fact-checked to identify false or misleading claims made since the first round of the Liberia 2023 general election. The data collection process involved gathering election-related fact checks from three Fact-checking organisations– DUBAWA, The Stage Media Liberia (TSM), and Local Voice Liberia (LVL), specifically focusing on materials about the candidates and the election. The three platforms were selected because of their reach and relevance in covering elections in the subregion of Liberia. The timeframe was limited to reports published between October 10, 2023, to November 17, 2023. The period was chosen because it was in the heat of the election timeline covering both the first round and run-off elections, a time when political actors were heavy on their mis/disinformation campaigns.

To be able to select the stories or data for the study, the researcher visited the websites of the three fact-checking institutions and picked stories that were published with “Liberia election 2023” as a prefix. Other considerations were the names of the political actors or institutions involved in the elections in Liberia. The stories were selected from October 10 to November 17, 2023. Where the platforms conducted and published the same fact-check on a claim, the researcher used it once in its analysis even though it acknowledged that report to have been cross-published by other platforms.

Through the content analysis, the researcher explored the themes and patterns in the various fact-check reports by the three platforms and teased out the personalities behind the spread of the disinformation campaigns.

5.0 DATA COLLECTION

5.1 Themes

The narratives used in spreading the mis/disinformation became the themes. DUBAWA identified and categorised these themes for this analysis. Apart from differentiating between unintentional errors (misinformation) and deliberate misinformation (disinformation), the report also prioritises the social media platform majorly used to spread such narratives.

Claim	Medium	Verdict	Disseminator	Fact-checker	Content	Subject
NEC’s official results are only on pink and not on white sheet	Facebook	False*	Martin Kollie	TSM	NEC is sharing fake results on White Sheet.	NEC (negative)
Monrovia’s city mayor, Jefferson Koijee, shared 25kg with MCC casual workers	Facebook	Mostly true	Martin Kollie	DUBAWA	Jefferson is engaging in electoral fraud.	Jefferson Koijee (negative)
Liberia shut down borders with Sierra Leone	Website	True	Sierra Loaded News	DUBAWA	The government shut down borders to ensure electoral safety.	NEC (positive)
Unidentified agents are scanning voters’ cards for US$30	Facebook	Misleading	Martin Kollie	TSM	Accuse Unity party’s opposition of manoeuvring the election process.	NEC (negative)
The election process was temporarily postponed at a county due to late arrival of materials.	Unknown	True	Unknown	DUBAWA	There was late delivery of voting materials due to some mishaps in transporting them.	NEC (negative)
NEC website was down on election day	Facebook	True	Martin Kollie	DUBAWA	Poor management of the result-viewing website.	NEC (negative)
Pre-marked ballots were discovered at a polling centre	Facebook	True	DECH Media	DUBAWA	Election materials were altered before the election process began	NEC (negative)
A fake polling centre was found at People Liberation Party’s headquarters	Facebook	False	Martin Kollie	DUBAWA	Accuse the election commission of election manipulation	NEC (negative)
‘Smart Election Results’ published polls data	Website	False	Smart System (parody)	LVL	Viral election results with no credibility.	NEC (negative)
DW Africa says Weah is leading at the poll	Website/ Facebook	True but rectified	DW Africa	TSM*	Weah got a false lead.	George Weah (positive)
Fumbah Armah registered twice	Facebook	True	Martin Kollie	DUBAWA	Election manipulation	NEC (negative)
CDC chairman, Mulbah Morlu, had a WhatsApp chat with unknown person about the election result	Facebook/ WhatsApp	Misleading	Unknown	LVL	Morlu manipulated the election with the NEC chairperson.	NEC (negative)
NEC used one party to represent two presidential candidates	Facebook	True	Raphael Johnson	TSM	Criticise NEC for the error.	NEC (negative)
The European Union (EU) has 90% of the results from the just-concluded poll.	Facebook	False	Truth Breakfast show	DUBAWA	Boakai is leading the poll as at publication time.	Boakai (positive)
Poverty is higher in Sierra Leone and Guinea than Liberia	Radio	Misleading	Frank Gibson	DUBAWA	Liberians should be grateful for Weah’s economic management.	George Weah (positive)
Unity Party’s Prince Toles won District 8 election	Facebook	True	James Samuel Goodday Flomo	TSM	A victory for Prince Toles and the Unity Party.	Unity Party (positive)
Costa’s Root FM was shut down for supporting and aiding Jestina Taylor	Radio	Misleading	Jestina Taylor	TSM	George Weah’s CDC oppressed opposition.	George Weah(negative)
Boakai Amara Kamara deleted video against Joseph Boakai	Facebook	True	Martin Kollie	TSM	Costa Show’s co-host is partisan against Boakai.	Boakai Kamara (negative)
Uncoordinated electoral activities was ongoing at a polling unit	Facebook	True	Martin Kollie	TSM	NEC could not manage the electoral process properly.	NEC (negative)
NEC chose an electoral officer (Aisha Bility) who supports the ruling party	Facebook	True	Martin Kollie	TSM*	NEC did not scrutinise their electoral officials.	NEC (negative)
NEC chose an electoral officer (Jacqueline Jackson) that is supporting the ruling party	Facebook	True	Martin Kollie	TSM*	NEC could not scrutinise their electoral officials.	NEC (negative)
Helicopter is not a means of transporting ballot materials by NEC	Radio	False	Henry Fomo	TSM	NEC used helicopters to transport ballot materials in previous elections due to deplorable road conditions	NEC(negative)
Boakai increased his security apparatus.	Facebook	Misleading	Nyantee Togba	TSM*	Boakai spent extravagantly on security	Joseph Boakai(negative)
UP Vice President stopped from voting at the wrong centre	Newspaper	Misleading	New Republic newspaper	TSM	New Republic newspaper mislead people with their incorrect headline	Unity Party (negative)
NEC has announced a winner as at November 15	Facebook	Misleading	James Samuel Goodday Flomo	TSM	NEC’s provisional result put the UP in the lead, but did not declare the UP winners a	NEC(negative)
UP’s Joseph Boakai releases a list of cabinet ministers, leaving out Nimba.	WhatsApp/ Facebook	Misleading	Unknown	TSM	The supposed chat displayed Boakai’s neglect of Nimba County.	Boakai (Negative)
CDC wants to sponsor attack on Nimba’s senator, Prince Johnson	WhatsApp/ Facebook	Misleading	Patrick Honnah	TSM	Spread fear about Prince Johnson’s safety.	CDC (negative)
Joseph Boakai told Voice of America that Muslims will not be appointed in his cabinet	Website	False	New Republic newspaper	TSM*	Mislead Muslim electorates to refrain Muslim electorates from voting in Boakai.	Boakai(negative)
Edward Appleton supported Joseph Boakai	Facebook	False	Unity Party social media group	TSM	Appleton wa accused of endorsing UP.	Unity Party (positive)
Weah said Costa worships him	Facebook	Misleading	James Samuel Goodday Flomo	TSM	Discredit Weah at endorsements ahead of run-off election.	George Weah(negative)
Finance Minister, Samuel Tweah want LRA to dismiss all Lofa and Nimba citizens in top position	Whatsapp	Misleading	Unknown	LVL	Discredit the finance minister for influencing the LRA against Lofa and Nimba citizens.	Samuel Tweah(negative)
Ghana’s last election had 10 million invalid votes recorded	Audio	False	Barsee Leo Kpangbai	LVL	Give credit to the integrity of NEC’s electoral conduct.	NEC(positive)
President Weah is training mercenaries to rig the elections	Facebook	Misleading	Evelyn Thompson	LVL	Discredit the president’s efforts to ensure a flawless election.	George Weah(negative)
President Weah is giving L$35,000 Grants for Campaign Support	WhatsApp	Misleading	Unknown	LVL	Praise the president for sharing money to electorates.	George Weah (positive)
Joseph Boakai’s Unity Party pays medical students’ stipends regularly	Facebook	False	Martin KN Kollie	LVL	Discredit George Weah’s administration for not prioritising payment of medical students’ stipends.	Boakai(positive)
Oscar Bloh of the ECC is partisan to the Unity Party	Facebook	False	Mannuel Fle Chae	LVL	Spread fear of possible electoral manipulation.	ECC(negative)
CDC orchestrated Erol Gwion and Alfred Dunner’s death	Facebook	False	Eugene Fahngon	DUBAWA	Accuse CDC of politicide and cultism.	CDC (negative)
NEC officials forgot about the Final Registration Roll at a polling unit and asked voters to simply write their names	Facebook	True	Martin Kollie	DUBAWA	Accuse NEC of incompetent officiating.	NEC (negative)

6.0 ANALYSIS

6.1 Fact-checking Organisations

In the analysis, as can be seen in the table above, DUBAWA contributed 11 publications that met the criteria, representing 28.9% of the total fact-checks. It also cross-published five claims (13.2%) with Stage Media Liberia. Apart from the cross-published reports, Stage Media Liberia had 14 publications aligned with the research objectives, accounting for 36.8% of the total fact-checks. Local Voice Liberia contributed eight publications, constituting 21.1% of the total fact-checks. These percentages from the 38 articles reviewed, provide insights and a detailed analysis of the kind of misinformation surrounding the election.

6.2 Typology of the claims

Out of the 38 fact-checks reviewed, there were 14 instances of true claims, outnumbering false claims (12), and misleading claims (12).

6.3 Medium Used

Also, it was observed that among the fact-checks conducted, Facebook accounted for 26 instances, while WhatsApp was the source of five claims. Other means of dissemination were Radio (three) and websites (four).

Facebook and WhatsApp, with their large user bases, were fertile grounds in Liberia’s information climate and were used for the rapid sharing of accurate and misleading information. The ease of forwarding messages on WhatsApp and the viral nature of content-sharing on Facebook may have contributed to the swift dissemination of misinformation, often reaching a vast audience before corrective measures could be implemented. This prevalence played a significant role in possibly shaping public perceptions and influencing voters’ behaviour during the election period in Liberia.

6.4 MAJOR THEMES

From the available fact-checks, the National Election Commission (NEC) emerged as a central subject of information flow, with a total of 18 claims circulating the commission. Strikingly, only two of these claims portrayed the NEC in a positive light. One of the claims was misleading, but the other was true. Most of the claims, numbering 16, were negative, with eight being false, some of which can be seen here, here, and here. 

In contrast, the information landscape surrounding key political figures was notably different. There were seven claims about George Weah during and after the election period, all of which were false or misleading. Interestingly, four of these claims presented Weah in a positive light, like this one. Meanwhile, his political party, the Congress for Democratic Change (CDC), had four claims, and about two portrayed the party negatively. On the opposing side, there were five claims about Boakai, none of which were true. While three claims presented Boakai positively, such as this, the remaining were negative. Boakai’s political party, Unity Party, faced four claims, with three being positive, like here, and one as negative. All these may have shaped public opinion about the candidates, their parties and the election’s outcome.

For clarity, claims marked “positive” indicate statements that present the subject in a favourable  light by emphasising the subject’s positive attributes, achievements, or actions. These claims contribute to the perception of the subject.

On the other hand, “negative claims” are assertions that portray the subject in an unfavourable  light. By emphasising the subject’s shortcomings or adverse actions, such claims set the subject up for criticism and negatively influence public opinion or perceptions.

6.5 MIS/DISINFORMATION FORMATS

The formats for the disinformation campaigns, whether text, pictures, or videos, give an idea about the sophisticated nature of the campaigns used during the election. According to a 2021 academic journal, “Seeing is believing: Is video modality more powerful in spreading Fake News via online messaging Apps?” it was found that video is causing individuals to perceive fake news as more credible than audio and text, and increases the likelihood of them spreading it.

The advent of AI-generated images and videos makes it even more complicated for voters to decipher what is true or false, so DUBAWA looked at the claims made in the fact-check reports under observation to determine the level of sophistication through the formats used during the Liberia election. We categorised these formats into seven, which are:

i. Text   – Text only

ii. Picture – Picture only

iii. Audio- Audio only

iv. Video -Video only

v. Text/Pictures – Text and Picture only

vi. Text/Video- Text and Video only

vii. Text/Picture/Video- Text, picture, and video

In interrogating the format used, DUBAWA analysed the 38 reports to ascertain which mode was used in the claims that were fact-checked. 13 of them were in a text, 17 had text and picture, four were in an audio format, two came in a video format and one for the combination of text and video, as well as text, video and picture forms.

7.0 DISCUSSION/SUMMARY OF FINDINGS

The study set out to explore and understand four major issues underpinning mis and disinformation campaigns. The researcher found rather curiously that the National Elections Commission, rather than the political candidates, was the biggest target of many disinformation campaigns, receiving 18 reports, 16 of which were negative in terms of treatment. Out of the 18, eight turned out to be false. The continuous dissemination of negative and false information about the NEC could have negatively influenced public trust in the electoral process and the commission’s credibility. In political theory and development practice, trust, mainly vertical trust (the belief the community members have in the institutions presiding over that community), is so crucial in the governance process that it cannot be downplayed. 

If the video mode is stronger and more impactful in disinformation campaigns, as espoused by Shyam Sandar et al, then it is fair to assume that little effort was used by political actors in doctoring videos for votes in the Liberia election. A total of 10.5 of the claims made had some video content. A total of 78.9% had text and picture formats.

The study showed that Facebook remained the largest vehicle used in spreading disinformation campaigns, which is a confirmation of Meltwater’s 2023 Digital Global overview for Liberia.  A total of 68.4% of the reports captured by DUBAWA were disseminated via Facebook. According to the Digital Overview in Liberia, Facebook is the largest social media platform used in the country, with over 623,000 users in early 2023.

While the study may not be exhaustive, it makes some  critical findings for actors against disinformation as a foundation for future research.

8.0 LIMITATION

While conducting this report, incomplete data posed a significant challenge, as the availability and access to misinformation were not uniform, potentially impacting the depth of the analysis, which could create gaps in the portrayal of the full scope and influence of false claims. Also, the analysis may not capture the entire landscape of misinformation, especially as certain platforms were less accessible or not included in the scope of the research.

Moreover, temporal constraints and underreporting of misinformation were critical limitations. The dynamic nature of misinformation and the rapid emergence of new false claims may affect the report’s timeliness and relevance. Delays in data collection and processing could affect the representation of the most recent developments in misinformation. Not all instances of misinformation are reported or fact-checked, and this poses the risk of underestimating the prevalence of misinformation during the election period. Therefore, DUBAWA recognises the potential gaps that these limitations may cause in this report.

The prevalence of false and negative information surrounding key political players and institutions raises questions about the role of misinformation in shaping the election narrative. Analysing the potential connections between this information flow and Boakai’s eventual victory is also challenging. However, it would have provided insights into the impact of misinformation on electoral outcomes and public perception. Also, Weah’s CDC leading the misinformation chart may have contributed to the party’s short-lived victory on October 10.

9.0 RECOMMENDATIONS

While this research provides some understanding about the formats of disinformation campaigns and the role of social media in the Liberia election, including which channels were used more, it is important to stress there is the need to do more to understand the extent to which some of these disinformation campaigns directly influence the outcome of elections. In the 2020 elections in Cote D’Ivoire there was evidence that disinformation campaigns were used to suppress votes by reducing turnouts there by influencing the outcome of elections. It is recommended that future research in Liberia elections will also look deeply at how disinformation directly or indirectly influences the outcome.

10. CONCLUSION

Social media played a significant role in the spread of misinformation during the 2023 Liberia general election. The proliferation of disinformation on social media platforms had a profound impact on the electoral process, as it created an environment in which it was difficult for voters to discern fact from fiction. This could have hurt the integrity of the election and the democratic process as a whole.

The findings of this report have implications for electoral integrity and information management in Liberia and beyond. Social media platforms can be used to spread misinformation and disinformation on a massive scale, and this has the potential to undermine the democratic process. Therefore, it is important to develop strategies to mitigate the spread of misinformation on social media platforms. This report provides a useful starting point for such efforts.