Rudransh Mukherjee, Shaily Desai, Dhruv Raghavan and Joyojeet Pal
14 August 2024Summary
With digital media competing with mainstream media to shape public opinion during election cycles, there are still open questions on the comparative value of different media platforms for the politicians’ outreach goals. We collected data on 151 politician interviews across 25 YouTube channels that interviewed Indian politicians in the last year and conducted a mixed methods analysis of the content and views of these interviews. We found that (a) political content garners comparatively higher engagement on channels that are primarily digital, than on YouTube channels of mainstream media outlets; and (b) there are significant differences in both the self-portrayal and the substantive engagement of politicians in interviews in the two mediums. Our findings underline the importance of a deeper examination of the importance of digital influencers, standalone anchors and journalists who set up their own social media channels to understand the challenges they pose to mainstream media houses.
Introduction
According to a report by DataReportal, as of the beginning of 2024, internet penetration in India stood at 52.4 per cent with 751.4 million internet users, of which 462 million have a social media presence.[1] Researchers found that digital mediums for electoral campaigning became an essential part of political strategy in 2019, with the Bhartiya Janata Party (BJP) spending increasingly more on digital advertising in the election cycle.[2] This growth of social media use has also led to the expansion of online news consumption, making social media an important space for negotiating political opinions. This has led to the creation of new classes of content creators, like news channels that operate solely online, which include independent journalists who leave media houses and operate their own short and long-form multimedia content channels, as well as digital-only news organisations (henceforth, digital media houses) that go beyond creating content on websites to maintaining a flow of material on platforms like YouTube. In parallel, we also see a massive expansion of digital influencers in various domains, including fashion, technology, comedy and food, who use their large online following to start engaging in news and political content.
We curated a dataset of interviews with politicians across YouTube channels, including digital media houses, independent journalists, digital influencers and mainstream media houses. We then mapped the timelines of these interviews, the rate of engagement by the public, and the style and content of the interviews to provide evidence for the broader questions around political engagement with online content. We also utilised the United States (US) and India Politicians Dataset[3] to analyse possible patterns in politicians’ retweets to media houses, influencers and journalists’ messages on X (formerly Twitter).
Context Setting
We examined four kinds of YouTube accounts for this study:
The results we are presenting here build on a significant body of recent work that has shown that political actors increasingly appear on a range of online spaces to do interviews, spreading beyond the traditional model of primarily engaging with professional journalists.[4] This trend has important implications for four reasons, First, there is a range in the ideological spread of various media sources, and politicians may find it useful to engage different channels depending on whom they are trying to reach. Second, social media affordances allow for advantages that are not available in traditional media, including engagement management, dovetailing into existing social media influence networks and micro-targeting based on factors such as age, region and political persuasion. Third, interviews on social media allow politicians to present themselves in informal, personable settings, without the constraints of time or space on television or print, without the unpredictability of live engagements, and with a potentially greater control of the policy narrative. Finally, interviews on social media also afford certain forms of adaptation that are better aligned with contemporary forms of media consumption – a typical social media channel on YouTube may offer shorts or clips of longer-form content, which can be recirculated on other forms of social media such as WhatsApp as clips. The infrastructure of social media content creation and circulation is equipped to handle video material with the intent of virality in ways that traditional media has not yet done effectively.
Politicians’ own adoption of these new forms of media is supported by changes in the broadcast news ecology. Recent work examining the role of digital influencers in political interviewing found that “unlike in engagements with mainstream media journalists, where a soft interview can undermine the credibility of a journalist, with influencers, this can be done while upholding the respective brand images of both parties.”[5] A digital influencer, without the training of a political journalist, may be far less informed about the fiscal, political or social implications of policy issues that a politician presents on their channel. This may end up endorsing a policy initiative by virtue of simply failing to ask the right questions and enabling the politician to control the general narrative.
Reach
Our data show that several of the highly influential mainstream media channels actually have considerably lesser engagement than channels with a smaller core following but a greater impact. To test if traditional media is being outperformed, we conducted a Mann-Whitney U-Test on the views received by politician interviews conducted by digital media and independent journalists (DMAIJ) versus traditional media houses. We find that DMAIJs have a combined average of 2,572 views per interview while mainstream media houses have 2,019, significant at .05.
Diving into specifics, we see that former NDTV news anchor Kumar emerges as the most engaged in terms of interviews, followed by Samdish and Curly Tales (the channel of lifestyle influencer Kamiya Jani), both of whom have a modest following compared to the major news channels but receive far more hits per video. While Kumar was already a well-known media personality from his work on mainstream television, Samdish and Jani largely owe their public renown to social media work.
Overall, the channel that does the most political interviews on YouTube (for the last four months of data) is Lallantop. The media outlet, which emerged a few years ago as an alternative Hindi language news and opinion source on YouTube, has grown exponentially in online popularity, eventually becoming part of the India Today and the Aaj Tak group. The channel has created a space for political interviewing, and in Saurabh Dwivedi, they have a prolific lead interviewer. The highest number of interviews overall are conducted by the YouTube channels of mainstream media organisations, followed by digital news organisations and digital influencers, with independent journalists getting the fewest political interviews. This suggests that the infrastructure of political interviewing, such as the networks that enable these, has extended into digital influencing.
We see a difference between the choice and content of politicians interviewed by independent journalists and digital influencers. Independent journalists’ YouTube channels tend to interview politicians around a current issue or controversy; thus, Kumar interviewed Satya Pal Malik over the liquor scandal, former television anchor and now independent YouTube journalist Faye D’Souza interviewed Aaditya Thackeray over the Mumbai racecourse controversy, whereas influencer Curly Tales’ interview of Thackeray’s rival Eknath Shinde has no particular theme other than to provide him with a lifestyle platform. Likewise, BeerBiceps enables a general conversation for politicians like India’s External Affairs Minister S Jaishankar to talk about a broad range of foreign policy issues. A look at the longer-term pattern of interviews on influencers’ channels shows that politicians show up on those as part of their regular public relations and outreach work during election and non-election cycles, while interviews with mainstream media tend to be more frequent closer to an election cycle.
Among digital influencers, on average, the highest median engagement accrues to interviews by Curly Tales, but that is very closely followed by Samdish and Raj Shamani. However, of these three, only Samdish does expert political interviewing, and this appears to be beneficial since he gets more median engagement than even BeerBiceps. Indeed, expertise does pay off; for instance, Technical Guruji, despite having a significantly higher following, receives far fewer engagements. Among the interviewers, Technical Guruji also has a style that seems most removed from the political style of back and forth, rarely challenging a politician during an interview. This suggests that rather than just the person conducting the interview, it is the content and style that matters in terms of the overall reach. When looking at the median engagement per video for the top three YouTube channels for media houses, independent journalists, digital media, and influencers, we see that the best-performing videos belong to Kumar, although his channel gets fewer subscribers than the others profiled. The most watched interview last year was with Jaishankar on BeerBiceps’ channel, which received over nine million views at the time of data collection in April 2024.
Interviews put out by mainstream media channels received fewer views on average than other channel categories. Many of these channels post upwards of 40 videos per day, thus, the attention put to amplifying individual content items may be relatively less, and videos can only go viral when some inherent quality in the content makes it so. However, many digital influencers only release one or two videos per week, and they put a great deal of energy into making these widely noticed through their outreach networks, including the use of short clips to draw in viewers. Given the smaller amount of content, they also have greater eyeballs per subscriber as compared to a larger media channel. It is also true that the standard terms of engagement that IT cells learnt from X do not apply to YouTube. You can AstroTurf content from your preferred politicians on X to turn it viral, but on YouTube, content is still king, and you need something engaging to attract and retain viewers; the name alone can only carry it so far.
Online Reach and the Value of Political Endorsement
What is the value that social media accounts get when a politician engages with them? Does it increase or decrease their net amount of engagement? To assess this, we turn to data from X to examine the value various accounts get from politicians. The results are apparent; journalists and digital influencers benefit incredibly when politicians retweet their content. More interestingly, media house channels fail to achieve the same level of political engagement as individual journalists or influencers.
We used the NivaDuck[6] and DISMISS[7] datasets to make a list of politicians, influencers, digital and mainstream media accounts and analyse their tweets from the US and India Politicians Dataset.[8] We map political engagements based on the National Democratic Alliance (NDA)/non-NDA on a single spectrum and attribute values to content and ultimately the account itself based on engagement by any known politician on either side. While this method puts all opposition parties into a single bucket, it nonetheless helps to understand the relationship of these accounts with the central government. We analysed tweets on X between 2020 to May 2023 (when X closed access to the public application programming interfaces) to compute these relationships.
We find that, on average, digital influencers and journalists’ content is 88 per cent and 74 per cent likely respectively to be more polarised in its consumption than both mainstream media and digital news channels.[9] Additionally, digital influencers are 73 per cent more likely to exhibit more polarity than journalists. We also find that as these accounts become increasingly polar, they tend to get a higher amount of activity on their tweets. This observation is supported by the Pearson correlation test between the polarity and retweet variables. Digital media has a stronger correlation of polarity with retweet value than mainstream media by a small margin. However, the same does not hold true for digital influencers and journalists. Regardless of whether influencers or journalists have a higher or lower median retweet value, all accounts are concentrated in the highly polar regions, which suggests that influencers and journalists inherently exhibit much higher polarisation than the other categories of X users.
A more evocative description of the impact of political engagement is seen when we compare the distributions of the retweet values from tweets engaged by politicians and those that are not. Our analysis of these distributions for mainstream media, digital media, journalists and influencers suggests that there is a staggering difference in the total scale of engagement between tweets that are engaged by politicians and those that are not.
While, in general, the expectation should be that politician-engaged messaging would have higher engagement values since the journalists, media houses, or influencers are themselves significant accounts, and only their least significant messages would be completely ignored by any politician since these accounts on average, have a fairly strong following. On average, for mainstream media, a political engagement gets 19 times the number of retweets received on any other tweet. This amplification factor for digital media is 18, for journalists is 26, while digital influencers witness a staggering 44-fold increase in their retweets on politically endorsed content.[10] This throughput advantage for digital influencers gives us a good indicator of why influencers see so much value in politicisation.
The Future of Politician Interviewing
Mainstream media outlets still dominated the total number of politician interviews, especially closer to the election. However, what is more interesting is that politicians are more actively engaged by digital influencers throughout the year. This means that the work of politicisation is an ongoing and year-long process, while the process of mainstream and digital media interviewing comes alive right at the end of the electoral cycle.[11]
Looking at only digital influencers, Samdish gets the highest amount of engagement. Although a few pairwise comparisons can be made for politicians who did interviews on more than one influencer channel, such as Nitin Gadkari, Piyush Goyal, Jaishankar and Smriti Irani, whose interviews notably got substantial viewership. High view counts for these ministers suggest their strong influence and public interest. In general, the interviews on BeerBiceps and Curly Tales do better than those on Technical Guruji and Kumar Shyam. We also note that overall, Samdish receives the greatest number of opposition-leaning politicians. Another channel that consistently gets good engagement despite fewer politicians getting interviewed on it is that of Shamani.
Interview Positioning Trends across Channels
We examined the differences between the style and substance of politician interviews on channels run by influencers, and journalists. We examined a case in which a politician gave an interview both on an influencer channel and with a professional journalist and reported our analysis.
Information and Broadcasting Minister Anurag Thakur appeared in an interview with food and lifestyle influencer Kamiya Jani on her channel, Curly Tales[12], as well as with journalist and anchor Sudhir Chaudhary on the India Today Group’s AajTak channel.[13]
Appearance: Beginning from his appearance in the two interviews, Thakur dons a kurta with a Nehru jacket, a BJP lapel and vermillion tika when being interviewed by Chaudhary from AajTak. On the other hand, when in conversation with Jani, he wears a nondescript blue kurta (Figure 1). The former has idiosyncrasies consistent with his political party while the latter appears more casual.
Atmosphere: Next, we see that the casual nature of the interview extends into the way the setting is used too (see Figure 2). In the interview on Curly Tales, Thakur is at a restaurant in Goa, and the camera spacing captures the casual atmosphere between the interviewer and interviewee; the table between the two and the items on the table allow space and comfort. The politician shows relatively eager body language, bending forward at times during the conversation to show relative closeness to the interviewer. The mainstream media interview, on the other hand, takes place in an office setting with the party logo clearly visible in the background, as well as statues of leaders that the party likes to lionise. The two men sit separated with their legs crossed in a defensive pose, and the tempo of the conversation remains formal and focused. The tone of both interviews is established in the first five minutes with both interviewers following a distinct approach to how they introduce Thakur. Jani begins the interview by calling him a “personality” and a “youth leader” from Himachal, with multi-varied interests and passions and welcomes the audience with a smile. Chaudhary, on the other hand, is much curter by referring to him as India’s “Information and Broadcasting Minister” and outlining that the interview is going to cover the BJP’s manifesto for the upcoming elections, given that polls were two weeks away.
Figure 1: A contrast of appearances between Anurag Thakur’s Interviews
Sources: Influencer Curly Tales (left) and Media House AajTak (right).
Figure 2: A contrast of body language and background settings
Sources: Curly Tales Interview (left) versus the AajTak Interview(right).
Language and conversational style also show certain distinctions. The topics the Curly Tales interview covers involve the personal life of Thakur and elucidate his interests, passions and hobbies growing up. This interview is advertised around Pahadi foods; thus, the viewer expects a cultural experience around Thakur’s home state of Himachal Pradesh. There are YouTube shorts spun out of the conversation which catch viewership of their own and circulates online for people who do not wish to listen to the entire interview. In contrast, the interview with Chaudhary has a serious air, he refers to political topics, has clear talking points, and has an aggressive style, pointing his fingers several times during the interview. Indian Prime Minister Narendra Modi is used differently in conversations across the two interviews. With Curly Tales, he attributes Modi’s hardworking nature in contributing to developing his own work ethic while he attributes his technical acumen in the AajTak interview penning an acronym for his surname by calling him the “Master of Digital Information” (MODI). The interview with a digital influencer presents Modi as a personal inspiration, with a guru-like tenor, whereas the interview with the mainstream media channels an argument for the political leadership qualities of the prime minister. There is also a difference in how Thakur speaks on behalf of the BJP in both the interviews. While there are allusions to the opposition in Jani’s interview, he explicitly targets the Aam Aadmi Party – specifically Arvind Kejriwal’s arrest – and accuses the party’s stalwarts of misleading the taxpayers of Delhi by engaging in the alleged liquor excise scam. The line of questioning also varies across both interviews, with Jani asking open-ended questions with the potential for Thakur to answer it how he pleases, while Chaudhary begins the interview with a much more objective line of questioning citing the United Nation’s (UN) statements against the BJP government asking him to justify the party’s stance against allegations of misusing the Central Bureau of Investigation and Enforcement Directorate.
Even though the AajTak YouTube channel has over 20 times the number of subscribers (over 60 million subscribers, compared to Curly Tales at about 3 million), the interview on Curly Tales gets over 10 times the number of views, holding 180k views, while the one on AajTak stands at just over 17,000 views.
Conclusion
In conclusion, we want to highlight a larger issue with the media ecology, outside of the politics of what it means for untrained digital influencers entering the socially critical space of political interviews. There are consequences for the future of traditional broadcast media, especially if we see the continued gradual decline of set-top-box-based television news subscription over YouTube-based news. In essence, the ‘Godi Media’ arguments may soon change from a traditional, digital binary to a battlefield that takes place solely on digital spaces like YouTube.
We could be experiencing the start of an overhaul of the media system in which influencers, indeed independent journalists or digital news channels with minimal budgets, are able to produce content that is consumed at higher levels than what mainstream media houses are able to do. This is extremely ominous for the future of news revenue since, at least in terms of net consumption, digital content from small producers already appears to be outperforming large media houses. While it is true that media houses offer more coverage, an increase in the skew in the monetisability of that content leads to a larger risk of a long-tail effect – a small amount of content being highly monetised, while a large amount of content gradually losing viewership and advertising revenue.
The repercussions of this are uncertain since there has not yet been any widespread impact of monetised subscription-based models for video content online. One immediate risk is of potential balkanisation of news into niche news, with the model of large newsrooms slowly dissipating. The success of breakaway channels such as those of Kumar, Anjum or Abhisar Sharma already point to the possibility that individual players can erode the revenues of mainstream media in the commentary space, and similarly, the success of digital influencers shows that they can likewise invade the political interviewing market. We may witness new strategies to regulate or control digital spaces like YouTube and other online channels that provide news since current trends point toward significant changes in the near future.
. . . . .
Dr Joyojeet Pal is an Associate Professor at the School of Information, University of Michigan. He can be contacted at joyojeet@umich.edu. Mr Rudransh Mukherjee is a student at Ashoka University and a former intern at the University of Michigan. He can be reached at rudransh.mukherjee_ug24@ashoka.edu.in. Ms Shaily Desai is a research intern at the University of Michigan. She can be contacted at shailyd@umich.edu. Mr Dhruv Raghavan is a student at Brown University and a former intern at the same university. He can be reached at dhruv_raghavan@brown.edu. The authors bear full responsibility for the facts cited and opinions expressed in this paper.
[1] Simon Kemp, “Digital 2024: India – DataReportal – Global Digital Insights”, DataReportal, 20 February 2024, https://datareportal.com/reports/digital-2024-india#:~:text=Here%20are%20 DataReportal’s%20 essential, penetration%20 stood%20at%2052.4%20 percent.
[2] Ronojoy Sen, Katharina Naumann and Vani Swarupa Murali, “The Impact of Digital Media on the 2019 Indian General Election”, Regional Programme Political Dialogue Asia, 19 December 2019, https://www.kas.de/en/web/politikdialog-asien/single-title/-/content/the-impact-of-digital-media-on-the-2019-indian-general-election.
[3] Anmol Panda, Libby Hemphill and Joyojeet Pal, “Politweets: Tweets of politicians, celebrities, news media, and influencers from India and the United States”, 26 May 2023, Social Media Archive, University of Michigan, https://doi.org/10.3886/xm68-rw44.
[4] Ronojoy Sen, Katharina Naumann and Vani Swarupa Murali, “The Impact of Digital Media on the 2019 Indian General Election”, Regional Programme Political Dialogue Asia, 19 December 2019, https://www.kas.de/en/web/politikdialog-asien/single-title/-/content/the-impact-of-digital-media-on-the-2019-indian-general-election; Nallu Preethi, Supriya Sharma, Siddharth Varadarajan and Vignesh Vellore, “How India’s Independent Media Are Collaborating to Cover the Parliamentary Elections…and Beyond”, International Journalism Festival, 20 April 2024, www.journalismfestival.com/programme/2024/how-indias-independent-media-are-collaborating-to-cover-the-parliamentary-electionsand-beyond; and
TNM Staff, “One More Gag Order: Eshwarappa’s Son Gets Injunction against 50 Media Houses in Bengaluru”, The News Minute, 1 May 2024, www.thenewsminute.com/karnataka/one-more-gag-order-eshwarappas-son-gets-injunction-against-50-media-houses-in-bengaluru.
[5] Sarah Khan, Rudransh Mukherjee and Joyojeet Pal, “Influencer collaboration on YouTube: Changing political outreach in the 2024 Indian Elections”, 28 January 2024, https://joyojeet.people.si.umich.edu/?p=1375.
[6] Anmol Panda, A’ndre Gonawela, Sreangsu Acharyya, Dibyendu Mishra, Mugdha Mohapatra, Ramgopal Chandrasekaran, and Joyojeet Pal, “Nivaduck-a scalable pipeline to build a database of political twitter handles for India and the United States”, In International Conference on Social Media and Society, 2020, pp. 200-209.
[7] Arshia Arya, Soham De, Dibyendu Mishra, Gazal Shekhawat, Ankur Sharma, Anmol Panda, Faisal Lalani et al., “DISMISS: Database of Indian Social Media Influencers on Twitter”, In Proceedings of the International AAAI Conference on Web and Social Media, vol. 16, 2022, pp. 1201-1207.
[8] GitHub Link to the Dataset: https://github.com/casmlab/politicians-tweets.
[9] These likelihoods come from the common language effect size after performing the Mann-Whitney U-Test between two distributions, that is, it is the probability that a value from Group A is greater (more polar, in this case) than a value from Group B. We conduct five tests with pairs for influencers and journalists together, and separately with mainstream and digital media, and report the average of two likelihoods where applicable. p-values for all tests range from 6.6e-16 to 0.0002, making them statistically significant.
[10] We report the median difference in magnitudes between retweet values for politician RT’d and other tweets from all accounts in the four sub-categories.
[11] We ran an ANOVA (analysis of variance) test with data from only the past 4 months and it concluded with a very low p-value (high statistical significance) that there is a significant difference between the mean interview dates for the channel categories. In addition, we ran t-tests to compare the mean data of the digital influencer category to the mean dates of the other categories. For mainstream news, independent news and digital media, we obtained small p-values, indicating strong evidence that the mean date for influencers is less than the mean dates for all media house categories. Overall, our analysis allows us to make the claim that media houses (mainstream news, independent news and digital media) conduct interviews closer to the election than digital influencers.
[12] “Sunday Brunch With Anurag Thakur X Kamiya Jani”, Curly Tales, 3 December 2023, YouTube video, https://www.youtube.com/watch?v=e9Zgly5QLLI.
[13] “Interview with Anurag Thakur, Interview by Sudhir Chowdhary,” Aaj Tak, 4 April 2024, YouTube video, 50:17, https://www.youtube.com/watch?v=q_Xkkqyd69c.
Pic Credit: Wikimedia commons