Social Media, Cyber Battalions, and Political Mobilisation in Ghana

Abstract

This report provides an overview of the mixed impact of social media on politics in Ghana. It draws on fieldwork carried out by the authors in Accra and Tamale between February and July 2019, as well as a custom survey in Tamale Central, Tamale North, Tamale South, and Nanton parliamentary constituencies. The report offers a set of policy recommendations for how to counteract some of the negatives and encourage the positives of social media for improving democratic accountability, facilitating intra- and cross-party linkages, and expanding voter outreach.

Full text

COLLEGE OF SOCIAL SCIENCES AND INTERNATIONAL STUDIES
Social Media, Cyber Battalions,
and Political Mobilisation in Ghana*
Elena Gadjanova, Gabrielle Lynch, Jason Reier, and Ghada Saibu1
*
The research for this report was funded by the Global Challenges Research Fund (GCRF) and
an impact fund award by the University of Exeter’s College of Social Sciences and International
Studies. We are grateful to Mumin Mutaru and the Tamale-based team of IPSOS-Ghana for research
assistance, and to Samantha Bradshaw, the Ghana Center for Democratic Development (CDD-
Ghana), the Institute for Democratic Governance (IDEG), and the Media Foundation for West Africa
(MFWA) for comments and advice on our research instruments. Most of all, we are grateful to all
of our research participants for their time, generosity, and insight. For further information on the
project, contact Elena Gadjanova (e.gadjanova@exeter.ac.uk).
1
Elena Gadjanova is a Lecturer (Assistant Professor) in Politics at the University of Exeter, Gabrielle
Lynch is a Professor of Comparative Politics at the University of Warwick, Jason Reier is a Professor
of Political Science at the University of Exeter, and Ghada Saibu is a Junior Fellow at the Bayreuth
International Graduate School of African Studies (BIGSAS) and a PhD Candidate at the University
of Bayreuth, Germany.
NOVEMBER 2019

2 3
CONTENTS
Summary 2
01. Methodology 4
02. Social media inuence is greater than
network penetration 5
03. Traditional and social media are not
separate spaces 7
04. Social and digital inequalities overlap 9
05. The big two political parties and
the market for digital entrepreneurs 11
06. The costs of increased investment
in social media 13
07. Misinformation is common 14
08. Some benets of social media 16
09. Conclusions 18
10. Recommendations 19
The use and abuse of social media in politics is a subject of increased interest around the world.
Social media has attracted much attention both for its capacity to facilitate mass organisation for
social change, and for its use to spread targeted, divisive, misleading, or overtly false content.
This report provides an overview of the mixed impact of
social media on politics in Ghana – a country that is often
held up as a shining example of democracy in Africa, with
its stable party system, closely-fought elections, and regular
peaceful transfers of power. Whether Ghana continues to
maintain this reputation depends in no small part on how
successfully the country navigates the challenges to democracy
in the digital age.
It is clear that social media is playing an increasingly important
role in Ghanaian electoral politics. Moreover, given the ways
in which messages and stories are shared and discussed, its
signicance is far greater than internet penetration gures
alone would suggest. The fact that elections – from party
nominations to presidential and parliamentary contests –
are often extremely closely fought also means that, while a
politician cannot rely on social media alone, an eective use
of platforms such as Facebook and WhatsApp can help to
make the dierence between winning and losing. Given the
(perceived) importance of social media to politics, there is
heavy investment in this space.
There are many positives to how social media is impacting politics
in the country. Politicians and their supporters are using social
media to inform and educate, to organise political meetings and
protests, and to mobilise support. These activities can ensure
that information is distributed more widely, encourage greater
and more informed participation, further boost voter turnout,
help to hold leaders to account, empower political activists,
and strengthen certain previously marginalised voices.
Yet, there are negatives as well. The fact that social media
is generally leveraged to bolster existing campaign strategies
is further exacerbating a number of problems. First, because
the use of social media is an addition to, rather than a
replacement of, older campaign strategies, it is further
increasing the cost of (already hugely expensive) campaigns.
The added expense only strengthens the position of
wealthy politicians and the two main political parties to
the disadvantage of the less well-o and smaller parties.
Second, because politicians and their supporters tend to
use social media to signal status through interactions with
constituents and inuential gures, as well as showcase
development activities, this further equates “good leadership”
with being (seen as) a “good patron”. While citizens can use
social media platforms to apply pressure on elected leaders, the
fact that such pressure often consists of demands to engage in
development projects may only bolster the existing patrimonial
logics of day-to-day politics.
At the same time, the use of social media to de-campaign
opponents ensures that rumours and misinformation, while far
from a new phenomenon, can now spread further and more
quickly. The ability to use fake accounts and pseudonyms, and
the closed nature of popular WhatsApp groups, also renders
such problems more dicult to tackle.
The digital age is also creating a new divide between rst-hand
and proxy social media consumers in Ghana, with far-reaching
implications for politics, representation, and social relations.
Inequalities in digital access and social media use exacerbate
socio-economic divides in the country and restrict the ability of
some citizens – most notably rural women – to exercise their
voice and to engage in politics.
Finally, there are other problems that, while not emphasised
by our interviewees at the time of the research, may be
hidden from view or become matters of concern in the future.
Examples include the use of big data to micro-target particular
groups with divisive messaging or misinformation, the employing
of spyware to hack ostensibly secure private messaging services
such as WhatsApp, and the ability of big business or foreign
powers to use social media to interfere with domestic political
processes and outcomes.
The rst section of this report outlines our methodology.
Sections 2 and 3 present a brief overview of social media
use and the relationship between social and traditional media
in Ghana. Section 4 addresses the eect of social media on
widening digital inequalities in the country. Sections 5 and 6
delve into the impacts of social media on politics and electoral
campaigns. Section 7 discusses misinformation. Section 8
presents a number of positives stemming from expanded
social media use and gives a number of reasons for cautious
optimism. Section 9 provides a brief summary and Section 10
puts forward a number of policy recommendations drawing on
the ndings from the research.
“Ultimately, any politician who tries to downplay or
underestimate the inuence of social media does so
at his own risk, because when things get o on social
media, it looks like the proverbial harmattan and dry
season. It is virtually uncontrollable.”
Political activist, Tamale, 27 June 2019.
SUMMARY

4 5
The analysis presented here draws on the authors’ existing knowledge of Ghana, as well as
65 in-depth qualitative interviews and eight focus group discussions with politicians, campaign
strategists, political communicators, political activists, youth group members, journalists and
civil society workers conducted between February and July 2019.
2 Fuseini, Issahaka, 2017. “City Prole: Tamale”, Cities 60: 64-74.
3 Ghana Statistical Service, “Urbanisation”, October, 2014, available at http://www.statsghana.gov.gh/gssmain/leUpload/pressrelease/
Urbanisation%20in%20Ghana.pdf
The research was carried out in the capital city, Accra, and
in Tamale and some of the surrounding areas. The report
also draws on a custom survey of 1,600 respondents in
Tamale Central, Tamale North, Tamale South, and Nanton
parliamentary constituencies designed by the authors and
conducted by IPSOS-Ghana in July 2019. All gures cited in
this report refer to results from that survey.
Accra was selected so that the authors could engage with
leading gures in the main political parties, civil society and
media. In Tamale and its surroundings, on the other hand,
the authors could study the dynamics of organization and
mobilisation on social media in an increasingly competitive
political space in a historically-marginalized part of the country.
Tamale is Ghana’s third largest city (population ~400’000 in
2013) and the second fastest growing metropolitan area in
the country.2 It is the capital of the Northern region, which is
among the poorest and least urbanized in Ghana with relatively
limited infrastructure and communication networks, and lower
literacy rates, particularly among women.3
In selecting Tamale and the surrounding constituencies as
a case study area, the research has sought to uncover how
citizens engage with social media beyond Ghana’s capital city
and in areas with relatively low digital literacy and internet
connectivity. To our knowledge, no existing studies have
addressed this question to date. The issue is central to tracing
the impact of growing social media use on politics and social
relations in the country, and to devising eective measures
to harnessing the positives and tackling the negatives of the
expanding digital space.
A second-hand smart phone can be bought for as little as GHS 60 (GBP 9 or USD 11) in the
markets of Accra and an increasing number of people around the country are active on social
media.4 The most popular platforms are Facebook and WhatsApp, with Instagram a distant
third. Twitter is also gaining ground but is generally regarded as an “elite” platform.
4Ghana is one of Africa’s largest mobile phone markets with close to 34 million subscriptions and over 10 million active internet users (about a
third of the country’s population).
01 02
METHODOLOGY SOCIAL MEDIA INFLUENCE
IS GREATER THAN NETWORK
PENETRATION
Figure 1. Which social media platform(s) do you use? Figure 2. On what device(s) do you access social media?
My own mobile
phone
On someone else’s
mobile phone
My own computer
At a public computer
On someone else’s
computer
Other

6 7
Given that the vast majority of stories aired on traditional media in Ghana now originate on social
media, no-one is effectively isolated from its reach.
This does not mean that social media content reaches
the broader public unltered, however. On the contrary,
journalists must strike a delicate balance when selecting
which social media messages to broadcast as they try to
gain a reputation for “breaking news” while simultaneously
trying to build and maintain a reputation as reliable and
trustworthy sources of information.
Traditional media sources enjoy relatively high levels of trust
in Northern Ghana. Radio and TV are seen as containing the
least amount of misinformation (Figure 4) and are trusted
more than political parties, state institutions, traditional
leaders, and even friends and family (Figure 5). Attention to
traditional media appears to be driven precisely by the desire
to hear from the “truth arbiters”, but maintaining this status
is treacherous.
The journalists we spoke to were keenly aware of the dangers
of jeopardizing their reputation and were overall committed
to maintaining it. Thus, they were selective of the social
media content they discussed, and the manner, in which they
discussed it. For example, Tamale’s radio hosts were proactive
in establishing cross-party WhatsApp groups and policing their
tone and content.
03 TRADITIONAL AND SOCIAL MEDIA
ARE NOT SEPARATE SPACES
Many who use social media check their accounts multiple
times a day and it is not uncommon for people to say that
it is the rst thing that they do when they wake up, and
the last thing that they do before they go to sleep. But not
everyone is online. Limited access to wi-, the high cost of
data (1GB of data costs USD 3.60, which is roughly equal to
two per cent of the monthly average income in Ghana),5 and
poor networks across much of the country further reduces
people’s access to, and use of, social media. This is particularly
true in rural areas.
5 Alliance for Aordable Internet, Mobile Broadband Pricing data for 2019, online at: https://a4ai.org/extra/mobile_broadband_pricing_usd-
2019Q2
Such limitations notwithstanding, the spread of social media is
far greater than what internet/smartphone penetration levels
alone would suggest. First, those who do have access to social
media expose others to online content: young people share
stories that they read with their families; clergy discuss news
in their places of worship; posts and tweets provide a popular
subject for debate in markets and other public spaces. This
spillover from online to oine social spaces is particularly
important given that friends and family are a key source of
news on politics and current aairs (Figure 3).
Figure 3. Where do you mostly get news on politics?
Second, debates on social media help to set the agenda
and to also shape the content of traditional media coverage
– from print journalism through to TV and radio. Journalists
we spoke to estimated that over 80 per cent of their leads for
stories were sourced from social media.
Figure 4. Which platform contains the most/least misinformation?

8 9
Digital inequalities matter for political participation, democratic representation,
and civic engagement. As politicians, state institutions, and journalists increasingly
seek to engage with citizens on social media, there is a real danger that inequalities
in data access and social media use will lead to inequalities in voice and representation.
Our survey data from Northern Ghana show that there
are dramatic dierences across the population when it
comes to monthly spending on data. Women spend less
on data per month than men (Figure 6), and there are
stark dierences in mobile data expenditure by education
(Figure 7), age (Figure 8), and place of residence (Figure 9).
At the same time, journalists and media houses need to attract
and retain an audience in an increasingly competitive media
space – as of July 2019 Tamale alone has 16 radio stations.
In such a context, journalists and media houses compete
for listeners and followers (and the all-important advertising
income that ows from the same). To this end, journalists solicit
listeners online by themselves using social media in a number
of ways. These include live-streaming programmes, advertising
or summarising programmes aired, collecting comments and
input from listeners, conducting online polls, and updating
listeners on “breaking news”.
Journalists’ extensive engagement with social media, together
with the fact that traditional media sources enjoy relatively
high levels of public trust, ensures that many social media users
select and circulate information from online news portals or
ocial media sites (Ghanaweb, Myjoyonline, and Citinewsroom
were the most frequently mentioned). There is thus a
constant information loop between social and traditional
media in Ghana.
However, the potentially positive nature of this loop is
undermined by the limited resources media houses currently
have for fact-checking, the little training that many journalists
and radio hosts can draw on, and the fact that the most
eective way to cultivate a loyal audience is to establish
a reputation for being (among the) rst to report on a
sensational story. All these factors in tandem can encourage
breakneck speed sensationalism, erode trust in traditional
media, and push fact-checking to the background.
A popular, but anonymous, social media account.
04 SOCIAL AND DIGITAL
INEQUALITIES OVERLAP
Figure 6. Monthly data expenditure by gender Figure 7. Monthly data expenditure by education
Table 1
Friends and family Other community
members
NDC NPP Other political
parties
Government
institutions
Traditional leaders Religious leaders Journalists and
radio hosts
Trust least 15.49 28.15 38.56 40.23 53.36 23.84 18.12 9.7 12.6
218.12 28.47 18.38 20.05 18.50 21.21 16.00 8.29 15.17
329.88 24.36 20.95 20.05 15.7 24.04 21.72 13.95 25.26
415.94 11.50 10.60 9.96 7.10 17.93 21.53 20.76 25.19
Trust most 20.57 7.52 11.50 9.7 5.34 12.98 22.62 47.30 21.79
Friends and family
Other community members
NDC
NPP
Other political parties
Government institutions
Traditional leaders
Religious leaders
Journalists and radio hosts
0
25
50
75 100
Trust least 2 3 4 Trust most
1
Friends and family
Other community
members
NDC
NPP
Other political
parties
Government
institutions
Traditional leaders
Religious leaders
Journalists and
radio hosts
Figure 5. How much do you trust information
coming from the following sources?

10 11
Ghana’s two main parties – the New Patriotic Party (NPP) and the National Democratic
Congress (NDC) – dominate the political space. In the 2016 elections they won over
98 per cent of the presidential vote and every single seat in parliament.
7 Authors’ interview with NPP communicator, Nanton, 29 June 2019.
Both parties invested heavily in social media for those elections
and their eorts have increased substantially since. Part of the
reason for this is a widespread feeling that the NPP’s more
eective use of social media in 2016 contributed to Nana
Akuo-Addo’s electoral success:
“Because [the NDC] couldn’t use social
media eectively, they did some projects
but it looked as if they did nothing [...] When
we noticed a little mistake from them, it was
what we hyped on social media. We used
[social media] extensively and it contributed
about 40% to our victory.”7
The NPP executed a collective and targeted social media
campaign strategy in the 2016 elections, and has since
maintained that structure. The party has at least one social
media communications ocer at every level of the party
structure (from the national oce down to the regional and
constituency oces). By mid-2019, they also had a “social
media army” of over 700 people on small allowances.
The fact that these “social media communicators” are
recruited on the basis of their vocal support for the NPP
online – as well as the other potential opportunities that
come from being a visibly eective party activist – also
encourages others (who are not on an allowance) to
actively promote the party and attack their opponents
on social media.
The NDC was slower to recognize the potential of social media
for electoral politics. The party’s 2016 social media campaign
was relatively disorganised and sporadic. Some individual
politicians – government ocials and members of parliament –
had small teams of communicators who promoted their works
on social media largely on a volunteer basis. The lack of a
centralized and coordinated social media team created internal
feuds between dierent social media factions within NDC who
rivalled each other for the attention of the president. This likely
contributed to confusion among supporters in 2016 and a
decrease in overall turn-out among those who had historically
voted for the party.
The NDC has now started integrating a coordinated social
media communication teams into the structure of the party
and is recruiting a large number of social media communicators,
but they still have some way to go before they catch up with
the NPP’s social media “machine”.
This uneven access prevents a large number of citizens from
fully beneting from what the internet can oer in terms of
information, advocacy, and political action. While access is not
fully prescribed for reasons described above, proxy access can
hamper citizens’ democratic participation and civic engagement
and entrench existing inequalities by forcing already marginalized
groups to rely on wealthier and more technologically-savvy
family members or friends to act as “translators” of online
content. Parties do attempt to create content that is accessible
to proxy consumers (mostly by relying on photos and audio
messages), but there is also a notion that proxy consumers are
more easily fooled, which can be exploited for political ends:
“Some [illiterate voters] will not be able to
read between the lines. Once they get the
information, they go with it. And before you
realize, a lot of damage is done. So, it was
very practical in 2016.”6
It is also clear that additional barriers to aordability, such as
the new nine per cent tax on communication services that
went into eect in October 2019, will compound existing
digital inequalities and will be disproportionately felt in rural
areas and among women, the elderly, and citizens with little or
no formal education.
05 THE BIG TWO POLITICAL
PARTIES AND THE MARKET
FOR DIGITAL ENTREPRENEURS
Figure 8. Monthly data expenditure by age group Figure 9. Monthly data expenditure by constituency
N.B: Tamale Central, North, and South are largely
urban constituencies, Nanton is rural.

12 13
This new digital marketplace creates a number of political challenges for the country’s smaller
parties (the CPP, PPP, PNC, APC, etc).
8 Authors’ interview with senior CPP communications strategist, Accra, 15 July 2019.
9 Focus group discussion with youth group members, Tamale, 26 June 2019.
These parties are acutely aware of the potential benets of
social media and are also investing more in this space. Social
media contributes to the survival of the smaller parties –
it oers certain advantages, such as relative ease in the
organisation of meetings and gatherings, and the maintenance
of a public prole.
However, social media is not currently closing the gap between
the “big two” and the smaller parties and might even be making
it worse. The limited resources of the smaller opposition
parties mean that they cannot invest in communications
teams or “social media armies” to the same extent, nor can
they generate the same level of voluntary engagement through
digital entrepreneurs. Thus, it is dicult to carve out a space
and to mobilise support. Because the NPP and NDC are able
to provide nancial and other resources to their social media
teams, the smaller parties are under pressure to do the same,
which compounds existing resource inequalities:
“If you don’t have resources, how do you
match them boot for boot? In 2016 we
tried our best to use social media, but
we were outnumbered, outresourced
by competitors.”8
And in a world where visibility is driven by numbers of
likes and followers, the messages of the smaller parties get
drowned out by the large and increasingly well-organised
social media machines and the armies of hopeful volunteers
of the big two. Finally, activists working for the smaller
parties also sometimes get put o by the ferocity of attacks
that they face online when they do criticise either of the
two main parties.
While important, social media has not replaced traditional
campaign activities – from donations and rallies to politicians’
attendance of weddings and funerals and participation in
discussions on the radio or TV. Having a strong social media
presence has become a necessary, but insucient condition for
eective campaigning. As a result, the need to invest in social
media constitutes additional work that has only increased the
cost of campaigns:
“You cannot take social media to
outdooring, you cannot take it to a
wedding. Even if you employ thousands
of people to manage your social media, it
is a must that you go to the outdooring.”9
The growing perception of the importance of social media
for politics and the two main parties’ willingness to devote
resources to campaigns online has created a market for digital
entrepreneurs in Ghana, much like in the rest of Sub-Saharan
Africa. Technologically-savvy operatives (most typically young
men) compete for the attention of local and national politicians
in the hope of being put on salary. Many volunteer to create
content promoting politicians (including setting up entire
Facebook, Instagram and Twitter accounts on their behalf ) and
growing their social media following. Existing youth groups are
repurposed to also operate in the new digital marketplace.
Not all entrepreneurial activity in the social media space
has partisan motives, however: our interviews revealed that
some social media activists seek to engage in a sort of digital
protection racket: they extort money from politicians by using
threats of circulating negative information (either true or false)
online unless they get paid. Thus, social media is creating new
opportunities for young people possessing a certain skillset,
but there is also a danger of spurring a new type of digital
political vigilantism.
Ghana’s political parties are not currently experimenting with
computational propaganda tools (using bots and algorithms) or
fundraising on social media, although there is certainly interest
in that direction. At the time of writing, there appears to be a
person behind every account and every comment/like, which
constrains virality and may limit the amount of circulating
misinformation. Relatively expensive data is a disincentive for
“click-bait” as the latter fails to generate the level of engagement
necessary to be protable. All of this may change with cheaper
data in the future.
Local party oce in Tamale.
06 THE COSTS OF INCREASED
INVESTMENT IN SOCIAL MEDIA

14 15
N.B: Categories are aggregated from open responses.
Interviewees and survey respondents agreed
misinformation on social media – or “fake
news” – was very common.
This ranges from outright lies to mischaracterizations,
exaggerations, spin, or purposeful ambiguity. Over half of our
respondents stated that they thought social media “sometimes”
or “often” contains misinformation and many could provide a
recent example of content they deemed fake news. We asked
if they could recall where they rst encountered recent fake
news and the responses paint a picture of how misinformation
transcends data and digital literacy boundaries in Northern
Ghana:
While the plurality of our respondents said they rst saw fake
news on Facebook, oine social networks (friends, family,
and community members) and radio were the main sources
of misinformation in rural areas and among people who did
not use mobile data and had little to no formal education
(Figure 10). Eorts to battle misinformation thus need to take
into account local dierences in how misinformation spreads.
In Ghana’s rural communities, these eorts should target
communities, not only individuals.
10 Focus group discussion with party communicators, Tamale, 27 April 2019.
11 Focus group discussion with youth group members, Tamale, 26 June 2019.
The range of fake news respondents cited is indicative of
what types of misinformation is conclusively debunked.
Examples include claims that are so outlandish so as to be
obviously implausible: rumours of persons being killed, going
missing, or being recovered/found alive, claims of completed
development projects later found to not have begun, rumours
of jobs and employment opportunities being oered later
found to be unavailable. These are all outright lies or strong
mischaracterizations, which are easy to fact-check and
show to be demonstratively false. But the subtler forms
of misinformation – exaggeration, spin, dicult to (dis-)
prove allegations, ambiguous statements – are conspicuously
absent from the list of conclusively debunked fake news our
respondents provided. Half-truths are obviously much harder
to dispel by producing concrete evidence.
Interviewees shared that misinformation is intended to serve
a number of purposes and is tailored accordingly: to gain
attention by becoming viral; to hurt an opponent; to counter
an attack; to divert attention from an important issue; to
confuse/”muddy the waters”; and to smear/delegitimise whistle-
blowers: “If you can’t convince, confuse”.10 Further, there is a
sense that – if the other side is believed to be using underhand
tactics – then it is also necessary to engage in the same given
the importance of winning. This encourages campaign eorts
that rely heavily on de-campaigning.
There is also a worrying tendency to create and spread
misinformation that could iname latent chieftaincy and religious
conicts in Ghana’s North. Chieftaincy issues get signicant
traction on social media because of their highly emotive nature,
so are an easy tool for those wishing to generate attention. In
the words of a focus group participant:
“[One thing that] trends most is chieftaincy
issues on social media. When you post [on
chieftaincy issues] the attacks and comments
are just coming. […] Because you are
attacking my personality. It is my personality
you are attacking. You are condemning it,
you are destroying it.”11
07 MISINFORMATION IS COMMON
Figure 10: Where did you rst see recent examples
of ‘fake news’?
All
Rural areas
Urban areas
Don’t use
mobile data
Spend more
than 40 GHC
No formal
education
Senior
Highschool
One of the conclusively debunked chieftaincy rumours circulating on social
media at the time of the research.
There may also be a more sinister motive behind such rumours – chieftaincy issues can be politicized in order to sow division
between ethnic communities for electoral gain. There is a history of such tactics being employed during election campaigns in
Ghana’s North12 and social media provides an easy to use new medium where inammatory content can be propagated while
hiding behind the anonymity of ghost accounts.
12 Steve Tonah, “The Politicisation of a Chieftaincy Conict,” Nordic Journal of African Studies 21, no. 1 (2012): 20–
Figure 11. What would you do if you suspect
something is ‘fake news’?
Ignore/
do nothing
Conrm from
family/friends
Investigate/do
research on
my own
Stop trusting
media/source
Report/inform/
comment
Blame/insult
source
Stop spreading
Once started, a rumour related to chieftaincy will
spread extremely quickly and widely: at the time of
the research for this project, a conclusively debunked
rumour about the death of the Dagbon king was the
most widely cited example of “fake news” respondents
had recently encountered. It had spread quickly and
gone beyond social media to reach communities living
in remote rural areas via phone conversations and
word of mouth.
Notwithstanding the level of Ghana’s democracy,
many respondents still think social media usage in
political activism can lead to political victimization in
the form of attacks and hurt the prospects for job
acquisition if the identity of the activists is known. This
fear leads many activists to create pseudo or “ghost”
accounts. Finally, when confronted with suspected
misinformation, only about half of our respondents
expressed willingness to seek conrmation, investigate
further, or report/raise awareness (Figure 11).
We nevertheless nd the responses encouraging. While
there is undoubtedly some social desirability bias here,
the responses reveal what are seen as appropriate
reactions to misleading content on social media, i.e.
a social norm around online communication. A little
under half of respondents see themselves as taking
action when faced with suspected misinformation.
But in Ghana’s highly competitive media and political
environment, turning o any supporters because of
a reputation for fake news can be very costly. Our
interviews suggest that media personalities, politicians,
and political activists are all aware of this danger, which
encourages caution and connes the most egregious
form of misinformation to shadow or ghost accounts.

16 17
A FOCUS ON ATTRACTING NEW
VOTERS HELPS TO MAINTAIN
A LEVEL OF CIVILITY IN SOCIAL
MEDIA DISCOURSE
Elections in Ghana are very closely-fought and parties compete
for the support of the so-called “oating voters” (undecideds)
and “entrants” (newly-enfranchised voters). Social media is
seen as a key tool for growing party memberships among
both groups and this goal inuences the type of content that is
communicated online:
“The aim of the party is to attract new
members. So, if your agenda is to insult
others, we will remove you from the
platform. It happens very often, we remove
people from the platform very often. […]”16
The focus on expanding reach and on winning over new
supporters through social media, combined with the belief that
people will be put o by personality attacks and dirty tactics,
creates strong pressures to maintain civility online:
“You have to ght […] and win the public mind in the social
media space. Sometimes, it gets so abusive. You have to pull
your troops back. [...] One thing I keep warning our teams
about is our sensibilities and sensitivities as a culture, and
don’t be pulled, or dragged into a debate space that oends
sensibilities. Otherwise, however valid your case, you would
have lost.”17
Notably, the sanctioning of uncivil discourse can also extend to
closed intra-party platforms on WhatsApp: “If you are used to
insulting people on social media, even your party members are
not so interested in your posts.”18
16 Focus group discussion with NPP and NDC party operatives, Nanton, 29 June 2019.
17 Authors’ interview with senior NPP communications strategist, Accra, 11 July 2019.
18 Focus group discussion with NPP and NDC party operatives, Nanton, 29 June 2019.
19 Focus group discussion with NPP and NDC party operatives, Nanton, 29 June 2019.
IMPROVING DEMOCRATIC
ACCOUNTABILITY
Ghana’s citizens can use social media to apply pressure on
their elected politicians and political aspirants, for example by
opposing things that politicians have done or are planning to
do. Such pressure includes the #StopThatChamber campaign,
which forced politicians to drop a plan for a new legislative
chamber in 2019. This campaign had both online and oine
dimensions as the hashtag activism was accompanied by
scenes of activists being forcibly ejected from parliament,
which were then aired and discussed on traditional and
social media alike.
Citizens can also use social media to demand that politicians
meet their manifesto pledges or that they deal with other
issues within a constituency:
“For example, when a politician comes
to Nanton and makes a promise […] And
then you vote for them and after a year,
two years in power that promise has not
been delivered. Especially when the promise
came with delivery timelines. And you vote
for such a political party, and the timeline
passed without delivery. Such a thing when
you put it on social media it is like hot
‘maha’!”19
In this vein, a number of politicians explained how they had
been forced to nd money for a particular development
project – such as a road or bridge improvement – in order
to silence popular criticism online.
We found several additional reasons for cautious optimism. There is a general disapproval
of divisive and violent tactics, which helps to mitigate the extent to which politicians promote
these directly with much of the more problematic messaging seemingly being posted and
shared by supporters.
13 Focus group discussion with NPP and NDC party operatives, Nanton, 29 June 2019.
14 A recent study on the role of WhatsApp in Nigerian elections reached a similar conclusion. See Jaime Hitchen, Idayat Hassan, Jonathan Fisher,
and Nic Cheeseman, “WhatsApp and Nigeria’s 2019 Elections: Mobilising the People, Protecting the Vote”, University of Birmingham, July 2019.
15 Authors’ interview with a youth group member, Tamale, 14 June 2019.
However, as a number of interviewees explained, the fact that
the politicians seem not to sanction such behaviour suggests
that they may be the ones directing or encouraging it.
A VIRTUOUS CYCLE
Most citizens, including people living in rural areas, feel that
news on social media cannot be trusted. This scepticism,
together with a willingness to seek further evidence or
conrmation, ensures an ongoing role for traditional media
and other actors. As already discussed, traditional media (radio,
TV, and newspapers) enjoy high levels of public trust and are
relatively eective “arbiters of truth”.
This trust helps to create a virtuous cycle whereby fact-checks
– particularly when they come from established news stations
– are often believed, and accounts shown to be spreading
misinformation are shunned, which in turn dis-incentivises the
open peddling of falsehoods: “When you are found posting
[misinformation], people become disinterested in your post.”13
EMPOWERING YOUNG ACTIVISTS
While the rise of social media has exacerbated certain
inequalities, it has also given new opportunities and power
to the predominately young men who work as social media
communicators. Some spoke of how this went hand-in-hand
with greater respect and inuence. However, it is important not
to overstate this change: activists still largely work for wealthy
politicians and depend on existing patron-client relations.14
FACILITATING LINKAGES
AND OUTREACH
Social media is also serving as a platform for intra and inter-
party networking. Political operatives can get to know others
from within their parties and from across the political divide.
Locally-based activists nd it easier to engage in political
debates at the national level. Grassroots and national-level
politicians are also able to interact and share ideas on group
platforms. This is a signicant departure from the past when
very few could network across party lines, largely by being
selected to appear on national radio or TV. Such networks are
often established and sustained in the hope of future benets
when power changes hands. Regardless of motive, however,
cross-party connections can facilitate political tolerance and
guard against polarisation in the long run.
In addition, social media has made it possible for politicians to
reach out to voters in what had previously been considered
“unsafe” or “no go” zones or communities:
“During the campaign there are some
politicians who cannot go to certain places.
He cannot go but he can send his message
to them through social media. You know, for
NDC and NPP there are camps. […] So,
for instance, if you are [from the opposition
party] and you want to do campaign in our
community, it might be dicult for you, but
if you post your message on social media,
we will read it.”15
Such outreach can loosen the grip of established party
networks and give citizens in non-competitive areas a political
alternative. It could also discourage political vigilantism, which
is largely sustained by the need to protect parties’ “vote banks”
largely by intimidating opponents.
08 SOME BENEFITS
OF SOCIAL MEDIA

18 19
■ Safeguarding trust in traditional media is paramount.
We recommend that journalists and media houses be
given support to fact check – for example, through less
commercialised ventures and targeted media trainings.
This support will help to guard against the impact of
misinformation and maintain the relatively high levels of
public trust in local and national media.
■ Local social hierarchies can be engaged in fact-checking as
well: for example, traditional and religious leaders often
enjoy high levels of trust in rural areas.
■ Encourage continuing scepticism of information on social
media, teach citizens to recognize and report ghost accounts
in particular.
■ Work to establish and strengthen a norm against passively
tolerating fake news. One potentially eective campaign
could see this tied to notions of civic duty.
■ Do not underestimate the communal aspect of the
spread of misinformation: interventions should also target
communities, not just individuals, particularly in rural areas
and among low literacy voters.
■ Improve digital literacy among women, in rural areas, and
among citizens with little or no formal education with a
view to more rst-hand and critical engagement with
online content.
■ Keep the costs of interventions low – data-intensive fact-
checking will be ineective.
■ Support smaller political parties to aid them in carving out
space for their messages on social media. Encourage radio
and TV hosts to feature communicators from the smaller
political parties on air and to actively recruit them as part
of cross-party social media groups.
■ On the basis of the ndings, we would advocate against
interventions that seek to overly police or close down the
social media space in Ghana, such as internet shutdowns
or the regulation of who is permitted to post blogs.
Such measures can easily be used to further strengthen
incumbents and dominant political voices.
■ It is important not to over-react to, or politicize, the need
to contain the spread of misinformation. Such moves will
be counterproductive and undermine trust in the very
institutions tasked with sustaining democracy. Citizens
should be trusted in their ability to lter out and sanction
misleading and inammatory content online, and equipped
with the skills and information needed to allow them to
eectively do so.
Notwithstanding the positive outlined above, it is important to re-iterate that social media
is decidedly a double-edged sword for politics in Ghana.
20 Kwami Ahiabenu, Gideon Ofosu-Peasah, and Jerry Sam, “Media Perspectives on Fake News in Ghana”, May 2018.
There are a number of potential dangers that demand
attention: rst, as we know from experience in other settings,
trust in traditional media is fragile. Safeguarding it requires a
proactive and coordinated response by individual journalists,
media houses, political operatives, state institutions, and
citizens. Ghana’s journalists are widely expected to identify
and call out misinformation, yet are woefully ill-equipped to do
so.20 Second, social media has become the medium of choice
for propagating fake news, including content that could iname
chieftaincy disputes in Northern Ghana.
Fake news originating online easily spreads oine through
tightly-knit communal networks and citizens with little to
no formal education are particularly vulnerable. Third, social
media compounds existing socio-economic inequalities in the
country and risks translating unequal digital access to lack
of political representation and voice. Fourth, social media is
making election campaigns even more expensive and increasing
the gap between the big two political parties and the smaller
opposition parties. In the process, while initially improving
parties’ capacity for mobilisation and outreach, social media
is also rearming, rather than supplanting, existing patronage
logics of politics in Ghana.
09 10
CONCLUSIONS RECOMMENDATIONS

20
2019SSIS051
CONTACT
Elena Gadjanova, e.gadjanova@exeter.ac.uk