Indian Journalists on Twitter and their Politician Footprint

Soham De, Joyojeet Pal

(to cite: De, Soham, and Pal, Joyojeet (2022) Indian Journalists on Twitter and their Politician Footprint. Accessed online at https://joyojeet.people.si.umich.edu/journalists)

Recent discussions around large corporations maintaining watchlists of journalists have added new questions around how the positions of journalists on issues can be understood vis-a-vis institutions they report on. In India, claims and counter-claims around partisanship in the media are ubiquitous, and both the media sources one consumes and the individual journalists one follows can be a marker of one’s political positions. 

Data & Methodology

To map out partisanship, we turn to politicians, and the journalists they engage as an indicator of whose journalism is broadcast by which party. We did this by using a (subset of the) database of approximately 40,000 Indian politicians from Nivaduck, and another database of 4099 Indian journalists from DISMISS, and examined every retweet (excluding quoted tweets) contained within the last 1000 messages (or less, if the politician tweeted less frequently) from politicians’ accounts dating backwards from September 14, 2022. The earliest message from a journalist retweeted by a politician in our data is from August 1, 2021.

To account for large imbalances in the number of politicians from every party in NivaDuck (the original dataset of politicians), we sample the top 200 politicians (by followers) from every major party (BJP, INC, AAP, SP, DMK, AIADMK). In our polarity analysis, we often group (AAP, SP, DMK, AIADMK) as Others. We therefore randomly sample an additional set of ~600 BJP and INC respectively politicians to normalize the number of politicians in each group. Finally, we omit all journalists who have been RTed for fewer than 10 times over the period of study to reduce the prevalence of outliers.

The classification of whether or not one is a journalist is done using the profile text data (ie if the individual notes that they are a journalist), and while we do weed out some accounts that are not active journalists, for the majority of accounts we rely on self-reporting. Also, we use “ruling party” synonymously with the BJP, and “opposition” with all other non-BJP parties. 

Network Analysis

We use Gephi – an open-source network analysis software to visualize politicians’ retweet networks. In the network visualization below, an edge represents a retweet of a journalists’ tweet by some politician in our dataset (NivaDuck). We use Force Atlas 2 (a graph layout algorithm) to visualize the network. Simply put, imagine the nodes to be like-charged particles repelling each other, and the edges to be elastic bands (higher the weight, stronger the pull) holding them together. Finally, we also apply an iteration of ‘Label Adjust’ on Gephi to improve visibility of the node labellings. 

Binary Polarity Score

We use a crude and simple indicator of polarity based on politicians’ retweets of a journalist. We define binary polarity (say A vs B polarity) as (Number of RTs from A – Number of RTs from B)/(Number of RTs from A + Number of RTs from B). This approach, while simple, suffers from several drawbacks and is not suitable for supporting strong claims. We refer the reader to Voting with the Stars: Analyzing Partisan Engagement between Celebrities and Politicians in India for a more nuanced and robust metric to measure political polarity.

Findings

Media House Polarization

We find that politicians from major parties tend to show preferences for certain publications. Three media groups fall solidly on the BJP side in terms of engagements by politicians – the Zee Group, the Republic group and the Jagran Group. On the opposition side, NDTVs Hindi and English offerings both tend to be engaged mainly by opposition (@ndtvindia by Samajwadi Party and @ndtv by INC), while the India Today group also generally falls on the opposition side. From among large-circulation broadsheets, @dainikbhaskar has a very high engagement by non-BJP politicians. 

Some media houses have media channels which fall differently on polarization. For instance, @TimesNow and @TimesofIndia lean BJP, in terms of the party handles most retweeting the channel, while its sister publication @TOIIndiaNews leans INC. Similarly @ABPNews is most engaged by BJP politicians, while its Kolkata-based print publication @ttindia is much more engaged by non-BJP politicians.

Figure 1: Political reception of Indian Media organization as inferred from politicians’ retweets. Color of the bubbles represent the affiliation of the majority of politicians who RTed the journalist.

We also see an important role played by small numbers of politicians who engage mainstream news aggressively. In this case, a small number of INC politicians have been extremely active in engaging the content on @dainikbhaskar. 3 hyper-tweeting politician handles – @rameshmeena63 (Sapotra MLA Ramesh Meena), @PSKhachariyawaS (Jaipur Civil Lines MLA and Minsiter of Transport Pratap Singh Khachariyawas) and @Premraj99443093 (an account since deleted/ purged by Twitter) aggressively tweet local rajasthan media orgs. The same handles are also important in engaging other news handles that publish a lot of Rajasthan-specific content including (@zeerajasthan_, @1stIndiaNews, @News18rajasthan). 

The majority of international publications tend to be engaged by opposition leaders, interestingly @CNNNews18 tends to lean heavily towards the BJP, whereas its content partner @CNN in general tends to lean liberal in the US.

Polarization of Journalists as inferred from politicians’ Retweets

When we turn to individual journalists, we find that the patterns are somewhat distinct. Unlike with media house handles, which are more evenly divided between the government and the opposition, individual journalists overall tend to be engaged much more significantly by the opposition.  As we see in Figure 2 below, the vast majority of journalist nodes (in purple) tend to be clustered around opposition politicians, especially the larger bubbles. 

The interpretation of the concentration of larger bubbles on the opposition side suggests that the journalists more retweeted by politicians tend to be on the opposition side, whereas the journalists with fewer individual tweets engaged by politicians tend to be on the BJP side. 

Figure 2: Visualized 8148 edges, 2174 nodes (smaller nodes are omitted for brevity) of the Retweet network of Politicians (from NivaDuck). Politicians are the colors orange (BJP), green (INC) and others (blue), while all journalists are purple. Size of the nodes are proportional to their total degree (indegree + outdegree). Edges represent politicians’ retweets.

The formation of the clusters on Figure 2  is oriented by party — the lower right cluster is BJP and the upper right cluster is mainly INC, AAP, and SP, with a few more opposition parties spread throughout). Perhaps one of the most important clustering facets is that most journalists tend to be preferred by one or another “block” – thus either retweeted by the BJP, or by a combination of INC, AAP, and SP. There are a small number of “crossover” journalists – the most significant is @pantlp, the editor of Dainik Bhaskar, who gets retweets from multiple parties. There are also a small number of “crossover” politicians, who retweet sources that their political rivals do as well — in Tamil Nadu we see this trend, where politicians from across parties tend to retweet a range of sources. For instance, both Sivaganga MP @kartipc and BJP state Vice President @vpalanisamy2010, who are prolific in their engagements with mainstream media articles, tend to use sources with a range of political positions. In part, this is also in part because some of the journalists or mainstream media channels that appear clearly on one side or another in national politics are not subject to the same lines of polarization in their stories’ significance for regional states.

One important trend we see is around journalists with large online following. We see that the BJP tends to have much stronger engagements with star influencers – ie high-followed anchor accounts in the 1 Million plus follower range — thus @RajatSharmaLive, @rubikaliyaquat, @DChaurasia2312 and @SushantBSinha each were retweeted over 150 times by BJP leaders in the study period alone.  In contrast, only two journalist with a mega following of 1 Million plus — @ppbajpai and @ajitanjum get more than 100 engagements from one or another major opposition party. 

Instead, what we see is that the journalists who get the most retweeted by opposition parties include the likes of @AadeshRawal @Supriya23bh @ranvijaylive @umashankarsingh @pantlp @puneetsinghlive @shakeelNBT – many of who are mostly political or party beat journalists or political editors with traditional newspapers. 

From among the high-following journalists, there are two other trends. First is that there tends to be overlap between INC, SP, and AAP in engaging some of these journalists, but rarely an intersection from both sides. From the accounts with over 1 million followers that had at least 50 engagements from politicians during the study period, not a single journalist had a moderate spread (ie less than 67% retweeting either for or against the BJP), suggesting a remarkably high polarization among the highest followed journalists. (See Table 2 for a full list of journalists whose retweeting from politicians is polarized) From the set of highly engaged journalists, the least polarized was Shiv Aroor, and even he scored almost 75% on the BJP side. Outside of these journalists with high politician engagement, only three highly-followed journalists had a relatively even spread of politicians across parties retweeting them among the high-followed journalists, they were @rahulkanwal, @shekhargupta, and @chitraaum.

When we break up engagement of journalists by party, and find that the Aam Aadmi Party politicians engage the most journalistic content both in terms of the number of journalists as well as the number of media handles retweeted per politician. BJP tends to engage fewer journalists, but tends to engage relatively more frequently per journalist. From all the major parties represented in Table 1, BJP politicians are least likely to engage mainstream journalists, for instance, the average BJP politician engages in fewer journalists per politician than INC or SP, but has a very comparable number of media channels engaged per politician.

Due to the scale of data, region-specific media are excluded from this step of the analysis, however, the party with the most journalist engagements outside of the four listed in Table 1 is DMK, with 495 journalist retweet engagements tracked during the study period.

Party of the PoliticianTotal Journalists Tweets RT’dTotal News Channels Tweets RT’dMedian (Mean) Number of Individual Journalists’ Tweets RT’d per politicianMedian (Mean) Number of Individual News Channels’ Tweets RT’d per PoliticianMedian (Mean) Number of Individual journalists RT’d per politicianMedian (Mean) Number of News Channels RT’d per politician
BJP588986224 (18.1)6 (18.4)2 (6.13)3 (5.5)
INC10164110185 (22.28)6 (20.94)4 (7.89)3 (5.83)
AAP2691323511 (19.5)10 (21.28)6 (9.38)6 (7.15)
SP2795162610 (26.61)4 (16.76)5 (8.28)3 (4.91)

Table 1: Number of journalists and media houses’ tweets engaged by politician and party

Overall, we see that a number of journalists whose content tends to be more engaged by opposition parties are either part of small media houses, or established journalists who have set up their own independent youtube channels. These include @ajitanjum @ppbajpai @abhisar_sharma and @sakshijoshii – most of these also do traditional on-the-ground reporting rather than anchoring. 

Figure 3: Top 25 Journalists (by follower counts) visualized for their inferred BJP polarity — here a 0.5 polarity on the x-axis means about 75% of the retweets they get from politicians belong to one side on the binary classification (in this case either BJP or non-BJP) 

We see in Figure 3 that the three most retweeted journalists from the high-following individuals are @ajitanjum on the opposition end, and @rubikaliyaquat @sushantbsinha and @rajatsharmalive on the BJP side. More importantly we see that very few journalists fall below the 75% threshold of inferred polarity, suggesting that at the top end, individual journalists’ content is rarely engaged across this binary party line.

Of the large number of journalists who tend to get engaged solely by pro- or anti-BJP, a number got absolutely no engagement from BJP-politicians in the study period including @sardesairajdeep, @Abhisar_Sharma, @RanaAyyub, and @SuhasiniH. However, most of the high-following journalists who were not retweeted by BJP tended to have engagement from both INC and AAP, and in some cases SP, from among the major parties. Rana Ayyub, for instance, has no retweets from BJP politicians, but is retweeted by each of INC, AAP, and SP in comparable proportions.

Figure 4: Top 25 Journalists (by politician engagement/ RTs) visualized for their inferred BJP to Opposition polarity

When we turn to the visualization fo the most engaged journalists, irrespective of the number of followers on Twitter, we see again that parties have their preferred journalists. The interesting pattern on the BJP side is that all the highly engaged journalists are also star journalists with large online following — the lowest-followed journalist with over 50 engagements from the BJP was @AmanChopra_ and he clocks in more than 330k followers.

On the other hand, the opposition side looks very different. The two most engaged journalists are @AadeshRawal and @Supriya23bh – both are political party beat journalists, and both have followings past 100k, but a number of journalists with relatively minor following of below 10k on Twitter, such as @shakeelNBT @siddharthjourno @ranjeetadadwal and @BabelePiyush are each very influential for their political reporting on opposition parties.

Commentators

We visualized the most influential commentators (Figure 5), who may or may not be full-time practicing journalists, who are highly visible in the media – either through bylines or through public commentary on various issues. These commentators are typically important for their annotations on specific issues rather than on-the-ground reporting, or anchoring live television. In this case, we see that there are way more commentators who are retweeted by BJP journalists than by other parties. So while @profdilipmandal and @_garrywalia are highly engaged on the congress side, there are several influential commentators on the BJP side, whose reach exceeds the 100k followers mark on Twitter. The two most important, by far, are @ARanganathan72 and @sgurumurthy, who get more political engagement than any field journalist or television anchor.

Figure 5: Top 25 Commentators (by politician engagement/ RTs) visualized for their inferred BJP to Opposition polarity 

Party Wise Analysis

We examined the most engaged journalists for four parties – given the size of the sample, we felt most comfortable presenting claims about four parties with a significant footprint in engaging the media – AAP, SP, INC, BJP.

Aam Aadmi Party

We visualized the journalists most retweeted by the Aam Aadmi Party, we find the top individual is @Isudan_Gadhvi, who was a journalist, then joined the party. Similarly, another journalist heavily engaged by the party was @anuragdhanda who also joined the party, as with @sharmanagendar, who also went from journalism to being an advisor to the Delhi CM office. Since these journalists were still part of the mainstream media ecosystem for some period of the study, they are listed here, but their role moving into the party system is important as it adds some mainstream media firepower to the party’s outreach system.

The most active political accounts of the AAP tweeting out journalists’ messages were @PulkitS_, @balakv1970 (former Infosys CFO) and @BajpaiDeepak.

Figure 6: Journalists most retweeted by politicians from the Aam Aadmi Party (with a minimum threshold of polarity of over 66% of the journalist’s total political engagement coming from that single party)

Samajwadi Party

The most active Samajwadi Party aligned accounts engaging with journalists’ content through retweets were @uddinizharss, @MohdFai05559506, @ManojSinghKAKA. Of these, Manoj Singh Kaka is a party spokesperson. The most engaged journalists, as visualised here in terms of the total volume of engagement is @WasimAkramTyagi followed by @pankajjha_ and @sengarlive. Interestingly, we see another former journalist turned political party operative – @pjkanojia, journalist who joined RLD, but has still received a great deal of engagement from Samajwadi Party cadres, in large part due to the poll alliance.

Figure 7: Journalists most retweeted by politicians from the Samajwadi Party (with a minimum threshold of polarity of over 66% of the journalist’s total political engagement coming from that single party)

INC

The INC has a fairly wide range of journalists that its politicians engage with, and as we see in figure, there are a fair number of relatively small bubbles (indicating that the journalist does not have a huge following). The politicians who most retweet journalists’ content from the INC include Bihar politician @Ahmad_Shakeel, MP @kartipc, and the party handle of the youth Congress in Punjab @IYCPunjab. In general, most parties do not have a major member of parliament engage journalistic content very significantly, but this is somewhat more common in Tamil Nadu. Besides INC’s @kartipc, @ptrmadurai, DMK Member of Parliament, is also very active in terms of retweeting journalists.

Figure 8: Journalists most retweeted by politicians from the INC (with a minimum threshold of polarity of over 66% of the journalist’s total political engagement coming from that single party)

BJP 

The handles belonging to BJP-affiliated politicians that most retweet journalists are @vpalanisamy2010 @balbirpunj @ManoJJoshiraj. Of these, only Balbir Punj is a former Rajya Sabha MP. As with other parties, very few of the high-ranking party officials retweet journalists. We see in Figure 9 that there are a number of well-recognized and highly followed anchors on this list, much more than any other party. 

Figure 9: Journalists most retweeted by politicians from the BJP (with a minimum threshold of polarity of over 66% of the journalist’s total political engagement coming from that single party)

Interactive Visualization: 

Here, we present a tool to play around with. You can visualize the level of polarity (based on the scores mentioned in the methodology) for any journalist who was retweeted by a politician during the study period. For best results, visualize a few journalists together to see how they compare.

Figure 4: Link to an interactive visualization (hosted on Streamlit) https://actuallysoham-twitter-journalists-data-visualization-app-17y3zp.streamlitapp.com/

Topic Distribution of Journalists’ Tweets that were RTed by politicians

We performed LDA to extract predominant topics from the set of all tweets authored by journalists which were retweeted by politicians. The goal was to identify the main themes that politicians choose to engage with.

Topic (s)Most representative* words in clusterInferred Theme
1minister, shri, birthday, spokesperson, senior, press, conferenceMinister/ senior politician related news
4, 8breaking, cbi, raid, addressing, live, gst, watchBreaking news
3, 6education, policy, budget, operationPolicy matters of general importance
5, 10aap, samajwadi, money, election, arrested, crore, lakh, courtElection and party related
7, 2, 9good, thank, happy, support, wishes, indian, scheme, share, congratulationsIn support of schemes, positive national sentiment

Most representative words are words that (a) almost exclusively appear in a given cluster and (b) non-trivial and (c) thematically represent most other salient words in the cluster. 

 Table 2: Highly Polarized Journalists (with over 90% engagement for or against  the ruling party)

usernameOccupationOrganizationFollowersPol TweetsLean BJPLean INCLean OtherLean Party
SinghPramod2784EditorThe New Indian14335401.00-1.00-1.00bjp
MinhazMerchantCommentatorIndependent275020351.00-1.00-1.00bjp
IyervvalCommentatorIndependent119362271.00-1.00-1.00bjp
ajitdattaCommentatorFrustrated Indian54720271.00-1.00-1.00bjp
prafullaketkarEditorOrganiser39141271.00-1.00-1.00bjp
prasannavishyChannel ExecutiveSwarajya131206261.00-1.00-1.00bjp
aditi_tyagiAnchorZee214121211.00-1.00-1.00bjp
AbhishBanerjCommentatorIndependent46094211.00-1.00-1.00bjp
DennisJesudasanParty/Political CorrespondentThe Hindu5543211.00-1.00-1.00bjp
UnSubtleDesiEditorOpIndia363929761.00-1.00-1.00bjp
mvmeetChannel ExecutivePostcard182013541.00-1.00-1.00bjp
shwwetapunjEditorIndia Today4208511.00-1.00-1.00bjp
hiteshshankarEditorPanchjanya9551491.00-1.00-1.00bjp
TrulyMonicaCommentatorIndependent67162411.00-1.00-1.00bjp
sgurumurthyCommentatorIndependent3115663120.99-1.00-0.99bjp
swati_gsCommentatorSwarajya2878422650.99-0.99-1.00bjp
SureshChavhankeAnchorSudarshan TV506678870.98-1.00-0.98bjp
abhijitmajumderEditorEarshot555186850.98-1.00-0.98bjp
AmanChopra_anchorNews182975252490.98-0.98-0.99bjp
sunandavashishtCommentatorEpilogue202654460.96-0.96-1.00bjp
DChaurasia2312anchorNews Nation12931121750.95-0.98-0.98bjp
akhileshsharma1Party/Political CorrespondentNDTV161836380.95-0.95-1.00bjp
RShivshankarAnchorTimesNow156924380.95-1.00-0.95bjp
RAJAGOPALAN1951CommentatorIndependent34108740.95-1.00-0.95bjp
vikasbhaAnchorABP1021731070.93-0.94-0.98bjp
gauravcsawantAnchorAaj Tak1435332220.91-0.91-1.00bjp
IMinakshiJoshiAnchorIndia TV4435331280.91-0.92-0.98bjp
ARanganathan72EditorSwarajya5326913220.90-0.93-0.98bjp
ShefVaidyaCommentatorSwarajya6214571330.89-0.91-0.98bjp
rubikaliyaquatanchorABP37409612610.88-0.90-0.98bjp
SushantBSinhaanchorTimesNow10306802360.86-0.90-0.97bjp
AMISHDEVGANAnchorNews18655727580.83-0.86-0.97bjp
sidhantCommentatorWION96368510.76-0.80-0.96bjp
RajatSharmaLiveanchorIndia TV90496401950.76-0.79-0.97bjp
payalmehta100EditorANI366941080.76-0.76-1.00bjp
AdityaRajKaulEditorTV93447001750.75-0.78-0.97bjp
jgopikrishnan70Party/Political CorrespondentThe Pioneer87832220.64-0.64-1.00bjp
pantlpEditorDainik Bhaskar251187322-0.160.13-0.97inc
sengarliveCommentatorIndependent10960642-0.57-0.900.48sp
TheSamirAbbasAnchorTV9 Bharatvarsh13075738-0.58-0.21-0.21inc
_pallavighoshEditorNews1822009562-0.770.68-0.90inc
UtkarshSingh_AnchorUnseen India / Independent YouTube Channel83322221-0.90-0.06-0.04inc
AshishSinghLIVEEditorABP9480875-0.920.81-0.89inc
ravishndtvAnchorNDTV2140842103-0.920.17-0.24inc
mkvenu1EditorThe Wire20762563-0.940.62-0.68inc
KotaNeelimaCommentatorIndependent3208043-0.950.91-0.95inc
WasiuddinSiddi1EditorSalar e Hind6100050-0.960.48-0.52inc
DeepakSEditorEditorDesh 24×75444768-0.970.38-0.41inc
ashutosh83BEditorSatya Hindi235558969-0.970.51-0.54inc
rohini_sghAnchorThe Wire810114238-0.97-0.01-0.02inc
puneetsinghliveParty/Political CorrespondentBolta Hindustan69000272-0.980.46-0.48inc
umashankarsinghEditorNDTV338794566-0.980.61-0.63inc
AadeshRawalParty/Political CorrespondentIndependent1096101182-0.980.91-0.93inc
bainjalCommentatorindependent453133133-0.980.35-0.37inc
ajitanjumAnchorIndependent YouTube Channel1218223414-0.990.57-0.59inc
ppbajpaiAnchorIndependent YouTube Channel2490473183-0.990.80-0.81inc
sakshijoshiiAnchorIndependent YouTube Channel881823375-0.990.47-0.48inc
ranvijayliveParty/Political CorrespondentIndiaSpend186361653-0.990.38-0.39inc
Supriya23bhParty/Political CorrespondentNews Tak1331941090-1.000.99-0.99inc
rupashreenandaParty/Political CorrespondentNews183250547-1.00-1.001.00aap
abhisar_sharmaAnchorIndependent YouTube Channel154165465-1.000.54-0.54inc
omthanviCommentatorIndependent108803109-1.000.56-0.56inc
PrashantTandyCommentatorIndependent879248-1.000.92-0.92inc
BabelePiyushCommentatorIndependent8164193-1.000.96-0.96inc
siddharthjournoParty/Political CorrespondentANI430862-1.000.97-0.97inc
shakeelNBTParty/Political CorrespondentNavbharat Times7975192-1.000.99-0.99inc
bhanwarmeghEditorShunyakal2175076-1.001.00-1.00inc
ranjeetadadwalParty/Political CorrespondentTimesNow697982-1.001.00-1.00inc


Posted

in

by

Tags: