Use of social media bots in elections: Appropriateness and impact on election processes

In October 2020, Tanzania will undergo general elections to elect the President and Members of the Parliament. There is already emerging speculation in the streets with regards to these forthcoming elections. Some think it will be ‘business as usual’, an argument that can be related to what is known in the political science literature as ‘incumbency advantage’ (Gelman and King 1990). Others think it will follow the unprecedented move of the 2019 local elections whereby the returning officers disqualified the majority of the opposition candidates.

Further, others think there are mighty changes ahead or at least a challenging electoral process, emanating both from within the ruling party and opposition parties. Our interest in this election is informed not only by those different views of what will happen, but with the advancement and increasing use of technology like social media in elections.

The use of technology in elections spans across various stages from mobilizing, organizing, campaigning, voting and counting. As we have discussed in our previous articles, technology has been a double-edged sword in electoral democracy and processes. It has had both positive and negative impacts.

Cheeseman et al. (2018) have shown how adoption of electoral technology has had an unfavorable impact on electoral processes in both the developed and developing world. The example of the 2017 elections in Kenya is valid in this case which involved fake news and misinformation based on people’s psychology as informed by data obtained from social media (Crabtree 2018).

Similarly, technology has had a disparaging influence on voters and outcomes with examples from the USA 2016 elections (Grover et al. 2018) as well as the 2016 EU referendum (Brexit) in the UK: as exemplified by the Cambridge Analytica scandal.

On the other side, technology has allowed and facilitated democratic processes including mobilization and organization of voters. It can also be utilized to predict outcome of the elections (Singh et al. 2020). It has made election campaigns relatively cheaper through the use of technology in reaching many people using less time. In 2020, it is fair to predict that technology will play a central role in various ways. For instance political parties, in specific Chadema, have announced its deployment of a digital strategy in its various operations in 2020.

In this article, we continue to focus on social media technology and its use in elections. This follows our previous articles which have explained, in general, different aspects of social media in politics including its use in political parties and also in democratic processes.

We have discussed, for example, how misinformation diffusion and fake news that spread through social media are detrimental. Today, we look at social media bots and underscore their appropriateness and impact on elections. Bots have already infested the political social media sphere of Tanzania. These bots are usually associated with certain ‘camps’ related to certain political parties.

For example, on Twitter, we have seen a number of complaints from political influencers with regards to bots, which have been associated with the ‘Lumumba camp’ (i.e. the CCM office in Dar-es-Salaam). In Jamii Forum, there exists ‘camps’ that are labeled based on political parties – ‘Lumumba camp’ and also ‘Ufipa camp’ (i.e. Chadema head office in Dar-es-Salaam).

But the question is, what are bots? ‘Bots’ stands for software robots. In relation to social media, we have social bots. Social bots are a computer algorithm that automatically produces content and interacts with humans on social media, trying to emulate and possibly alter their behavior (Ferrara et al. 2016).

In light of this, bots can be positively useful in particular by assisting to pass or provide useful services or information. Thus, bots can be appropriately deployed with the right intention.

On the other hand, bots can be used for harmful purposes such as passing and promoting fake news, polarizing people in the same communities and using abusive language and distorting public information. Without a real human identity, bots can be used to carry out such activities on social media platforms.

As noted earlier, in Tanzania bots have been instrumental in cementing polarization based on political party association. They have been noted for abusing, making noise, promoting hate and misinformation. Political influencers in Tanzania have, on a number of occasions, tried to underscore the presence of bots, exposing them as well as educating their followers on how to identify bots in the context of Tanzania. The following quotes (tweets) are an example of those efforts in dealing with bots from a number of political influencers:

“Followers laki moja! Wow! Sikutegemea. Nawashukuru wote waliomua kunifollow (kwa kweli sijui kwa nini) na hasa waliozidi kutoa maoni, na ubishi, na mawazo mbadala maana naamini tunaelimishana hivyo. Bila kusahau mabot wenye matusi kibao. Bila wao nisingefikia 100,000.” – Richard Mabala; 14:06; 19/01/2020 (emphasis added)

Huu ujumbe ni kwa ma BOTS. Kama hamutaka kujulikana basi at least andikeni Tweets na msiwe mnatuma majibu tu tena kwa watu Fulani. For goodness sake, do you know NOTHING at all? Pili: mkiwa na followers 2 lakini mnafollow watu Fulani tu….Dead give away. – Fatma Karume aka Shangazi 10:10; 29/10/2018 (emphasis added)

Naona bots zimechanganyikiwa hawana hoja leo siku nzima Twitter na Instagram wanatukana matusi tena makali mpk wanatukana mama zetu. Sasa ushamba wao hawajui matusi = block so itabidi watafute akaunti zingine…wanajiongezea kazi. Just alert me hawa jamaa niwape tofali! – Maria Sarungi Tseshai 3:21; 18/12/2018. (emphasis added)

Ma ROBOTs wa Lumumba. Nothing Upstairs – Fatma Karume aka Shangazi 10:31; 29/10/2018 (emphasis added)

Umerealize Bots huwa wanaenda offline wote then after sometime wanarudi wote ka nyuki. Nowadays Those guys depend on Lumumba Wi-Fi. Connection ikiwa weak wanapotea wote. #7KNoMore - David @NotYetUhuru 3:46; 28/9/2018 (emphasis added)

The qualitative analysis of interpretation of ‘who are the bots’ in Tanzania from the Twitter sphere as well as being ‘followed’ by some indicate that the ‘bots’ in Tanzania are not software robots but actual human accounts that use fake IDs and are registered with a purpose of attacking or supporting a certain person or agenda, respectively. Thus the practical interpretation of bots in Tanzania is mostly based on its use more than its identity or technical formation.

Since the bots in Tanzania are ‘real humans’, it raises more concern with regards to the appropriateness and the implications to society. This also highlights the level of technological advancement that is evidently lagging – or the level of technological manipulation that is rising - which involves the use of human agency rather than algorithms and other technological tools that can facilitate software-developed bots.

In the 2020 elections, there is little doubt that there will be more of these human ‘bots’ that will be deployed and used to spread fake news and misinformation, abuse people, attacking, blackmailing and threatening people. There is real potential to deceive users. This is dangerous since it could translate to physical abuse. Bots, it can be said, have become increasingly sophisticated and harder to detect (Ferrara et al. 2016). Subsequently, the combination of ‘real’ bots and ‘human’ bots can result in higher potential for manipulation of citizens’ perceptions and citizens’ attitudes when it comes to relating to political content on social media platforms. This, in turn, can influence their voting behavior once the elections do arrive.