07

Disinformation on messenger services worldwide: A study of India

Indian case study

Infrastructure

To understand the online spread of mis- and disinformation and its implications for society in India, we first need to look at why and how the use of WhatsApp became so widespread in India. India is a country with a relatively low literacy rate (66 percent for women, 82 percent for men as of 201834). Sending voice messages and images via messenger services such as WhatsApp thus enables new forms of communication for many people. Decreasing mobile data prices have contributed to the use of these formats. Mobile phone use has been on the increase in India since 2003, and particularly the use of smartphones has risen sharply since 2013. The high tariffs of mobile internet were due to the structure of the telecommunications market and the high costs of 3G and 4G licenses. Network expansion largely focussed on urban areas. Even today, rural areas often only have 2G coverage, which is adequate for text messaging. The provider Jio, which is owned by India‘s largest private company, Reliance Industries, entered the market in 2016. The prevalence of Jio phones in India is in part due to the company‘s skirting of antitrust regulations, which was benevolently ignored by the government. For instance, towards the end of 2016, Jio was allowed to offer free phone calls and unlimited data for almost a year. This led to a restructuring of the telecommunications market and lower mobile data prices. Consequently, the telecommunications market in India is effectively an oligopoly, since other providers did not have the financial resources to burden the losses that Jio‘s offer created. In addition to significantly cheaper data, the cost of smartphones has also dropped by 16 percent over the last ten years34.

News consumption

India is a mobile-first market – even for news, where the primary device is a smartphone and not a laptop or desktop computer, as is the case in many Western countries. According to a Reuters study of the Indian market, 68 percent35 of English-speaking Indians primarily use their smartphone to access news online. The figures are significantly higher than in other comparable markets such as Turkey or Brazil. Of the 75 percent of English-speaking Indians who use Facebook, 52 percent get their news from it. 82 percent of English-speaking Indians use WhatsApp and of that 52 percent use the messenger for news. By contrast, news consumption through Instagram (26 percent), Twitter (18 percent) and Facebook Messenger (16 percent) is relatively low. 50 percent of those surveyed by Reuters share and/or comment on news – largely on Facebook or through WhatsApp. Almost as many (49 percent) were concerned that expressing their views publicly would change their friends‘ and families‘ opinion of them. 50 percent had the same concerns about the opinions of their colleagues and acquaintances. 55 percent are afraid that sharing their opinion may get them in trouble with authorities. Trust in news is low – according to Reuters just 36 percent of English-speaking Indians consider news generally trustworthy. At 39 percent, this figure doesn‘t improve much for first-hand news. Interestingly, 45 percent trust news from a search engine and 34 percent trust news from social media. Political affiliation doesn‘t change how much trust is afforded to news, and is even lower for people who feel they don‘t belong to a party and don‘t place themselves on the political spectrum. 57 percent of those surveyed question whether the news they receive has been manipulated.

Dr Shakuntala Banaji and Ram Bhat surveyed the non-English speaking population and analysed the spread of mis- and disinformation in India36, where it has been responsible for physical assaults, rapes and lynchings. The two researchers and their team focussed on the sociological and political aspects of the spread of mis- and disinformation. They found that gender has a significant impact on the use of WhatsApp, as well as how content is shared via WhatsApp. There is a substantial difference between the genders in terms of access to smartphones, private smartphone use, media literacy and the use of resources – from electricity to data. Women in India are disproportionately under threat of physical and virtual violence – increasingly so if they belong to a minority group including Muslim, Christian, Dalit or Adivasi. The research team found the most likely culprits for spreading misinformation, disinformation and hate news to be young or middle-aged male, tech-savvy Hindus of the upper or middle castes. They are also the most likely to create and manage groups on WhatsApp where mis- and disinformation is spread. If someone is part of a low caste, Dalit, Muslim, a woman, lives in rural areas and less tech-savvy, they‘re significantly less likely to create or manage such groups. Internet shutdowns by the Indian government, ostensibly to combat the spread of mis- and disinformation instead have the opposite effect. Limited access to news facilitates the spread of mis- and disinformation because it removes simple safeguards such as fact-checking information using a search engine. In 2019, the Indian internet was shut down 121 times. Venezuela, which takes second place for most internet shutdowns, managed just 12 shutdowns in the same period37. Amongst the Indian internet shutdowns of 2019 is the longest ever recorded worldwide: 213 days.

Misinformation and disinformation in India

In India spreading mis- and disinformation is used to discredit certain groups, particularly affecting marginalised groups fighting for justice, equality and democratic processes. Mis- and disinformation is also employed to normalise and legitimise discrimination and to cement the positions of socially dominant groups. A connection between the emergence of mob violence and WhatsApp use is evident. Mob attacks have predominantly occurred in states where the Hindu nationalist Bharatiya Janata Party (BJP) is in power and 86 percent of the victims of mob violence are men. In Hinduism, cows are considered sacred and cannot be killed. Most attacks happen because of this edict. Another common reason is the spread of rumours concerning strangers kidnapping children and/or trafficking organs. There is a clear correlation between the caste and religion of both victims and attackers. Messaging services have played at least one of three roles in acts of mob violence: 1. The spread of information about victims prior to the attacks. 2. The quick mobilisation of a lynch mob. 3.The spread of images and videos by the attackers or onlookers after the attacks. The spread of imagery documenting the attacks often leads to further violence, fear and tension. Banaji and Bhat identified five different categories for classifying the spread of mis- and disinformation in India: shocking content such as accidents, corpses, natural disasters and violence; nationalism and ethno-religious bigotry; religion; gender; and miscellaneous content – which includes content that did not incite violence but was shared by people whose content did at some point result in violence. The importance of the last category should not be underestimated. It creates a noisefloor: everyday content such as animal videos, recipes or videos of children singing, which can trigger positive emotions in the members of a group and creates a sense of belonging. The breadth of the effects of mis- and disinformation emerged from numerous interviews conducted as part of the study, including a decline in trust of traditional media. Interviewees believe that traditional media outlets deprive them of ‚real news‘ with WhatsApp and other social networks seen as the only source for ‚real news‘. As a result of disinformation, militant nationalism is bolstered and indirect threats against challenging the hegemony of the ruling class are issued. Using disinformation, a (Hindu) victim narrative was constructed, primarily based on conspiracy theories that vilify Muslims. An us vs. them narrative strengthens a sense of belonging for religious groups. Strongly gendered, pornographic, violent and voyeuristic content, is used to ridicule and discredit women and elicit general anxiety, fear of technology and self-censorship alongside causing depression, which can lead to suicide. Banaji and Bhat explicitly state that WhatsApp and other social media platforms are not the cause of mob violence triggered by mis- and disinformation. Nevertheless, social media networks and especially WhatsApp play a crucial role, enabling the rapid dissemination of content that incites mob violence. According to the authors, deep-rooted issues of inequality, prejudices of religion, caste and origin, as well as racism, misogyny and all forms of propaganda lie at the root of the problem and need to be eliminated. Regulating technology alone is not a solution; the previously mentioned internet shutdowns only exacerbate the problem.

Image disseminated on WhatsApp, which shows Nationalism and sexual violence combined. After the Suicide bombing in Pulawa, Kashmir, at the begining of the year 2019, this picture was widely spread. The text on the picture can be translated roughly as "With an oath on your mother, direct the Keep-going-Victory for Hindustan!"

Motivations for sharing mis- and disinformation

People often unwittingly share mis- and disinformation. As in other places, visual content is often viewed and shared uncritically in India. According to Banaji and Bhat, older people in particular are likely to believe and share information they receive from people in their community seen as trustworthy. It appears they see it as their duty to look after the community and alert others to threats, which may explain their behaviour. The social status of the initial person who sends a message plays an important role: the better their reputation, the more likely it is that untrustworthy information is taken seriously and spread. People are keen to earn a reputation as respectable and trustworthy by being the first to acquire and share information about local events. They position themselves as amateur reporters who provide ‘news‘ via WhatsApp. As a consequence of the race to be first, people often use existing imagery suggestive of depicting an incident when it, in fact, does not. A commonly expressed belief in the interviews held that television would not report on local incidents in order to cover them up. According to the authors, this reveals a need for more local journalism. Local journalists who know and understand the community would reduce the need for amateur reporters, who try to fulfil a need for local news but rarely work according to journalistic standards38.

33 Cf. Statista 2019a. 34 Cf. Banaji and Bhat 2019. 35 Cf. Aneez et al. 2019. 36 Cf. Banaji and Bhat 2019. 37 Cf. Mohan 2020. 38 Cf. Banaji and Bhat 2019.