https://orcid.org/0000-0001-5933-1283
Abstract
Scientific issues, as documented extensively for the Global North, are increasingly being subjected to various levels of politicization and polarization, often through popular media, in order to influence public opinion that is a key to democratic policy-making. In this article, we have chosen India, a functioning democracy in the Global South and the largest democracy in the world with a very diverse political landscape, to find out the extent of politicization and polarization in news media reports on COVID-19 and related issues. We have considered more than 200000 news articles from 6 Indian newspapers from 2020 to 2022 to measure the politicization and polarization of the issues based on computer-assisted content analysis methods. We have observed a significant and fluctuating trend of politicization of news coverage and polarization among two major political blocs of India over the 2-year time period of the COVID-19 pandemic, indicating the presence of differentiated politicization and punctuated polarization regarding the pandemic. We have also observed that, during this period, increasing levels of politicization were not necessarily accompanied by elevated levels of polarization in India. However, in the long run, the major political parties in India demonstrate a consistent level of polarization.
Introduction
The COVID-19 pandemic has affected the world in many ways. Some of the effects are the large-scale cleavages in the outlook and opinion that we have observed regarding the legitimacy of the pandemic (Boxell, Conway, Druckman, & Gentzkow, 2020; Flores et al., 2022), the legitimacy of the vaccines that were developed (Hotez et al., 2021; Ward et al., 2020), and the legitimacy of the precautionary measures (Pascual-Ferrá, Alperstein, Barnett, & Rimal, 2021) executed by various governments across the world. We have numerous examples of politicians, faith leaders, and other authority figures acclaiming dubious remedies, conspiracy theories, and unscientific means to tackle the spread of COVID-19 (The Guardian, 2020). Such debates and polarized positions are well documented for WEIRD (Western, Educated, Industrialized, Rich, and Democratic) countries. However, for non-WEIRD countries, scholarly efforts to observe, document, and analyze such polarization are scarce.
With about one-sixth of the world’s population, India is a significant member of the global south. It has a functioning democracy for decades. India has a high literacy rate of 96% among the youth and a gross graduation ratio of 31% (UNESCO, 2022). Since public opinion plays a decisive role in a functioning democracy (Dalton, 2013; Soroka & Wlezien, 2010) and the manner in which scientific news are presented to the public has a marked potential to influence the public opinion (Mihelj, Kondor, & Štětka, 2022), the politicization and polarization of scientific news is of remarkable consequence in a democracy like India. India has been home to the second-largest number of COVID-19 cases and the third-largest number of COVID-19 deaths (WHO, 2023). India is, therefore, an appropriate test bed among the non-WEIRD countries to assess the politicization and polarization of science in news media regarding COVID-19 and related issues.
During the crisis of COVID-19, there were a number of events indicating the politicization of the issue and polarization of the Indian society. On one hand, we have a Member of Parliament of the ruling Bharatiya Janata Party (BJP) denying the need for antiviral drugs with the words that “I consumed cow urine daily and it is a kind of acid which purifies my body. It also purifies the lungs and saves me from Covid-19 infection. I don’t take any medicine against corona but I am safe,” (Hindustan Times, 2021). On the other hand, an opposition party’s leader said that he would not take “BJP’s vaccine” (The Indian Express, 2021). Public opinion took an ugly religious turn, from the spread of Islamophobic hashtag of “CoronaJihad” targeting the entire Muslim community to spread COVID-19 (Perrigo, 2020) to an Islamic cleric announcing “Muslims must wait for fatwa before taking COVID-19 vaccine jab” (TimesNow, 2020). Krishnan (2021) mentions about the rumors about vaccines in India, of which one rumor particularly stands out saying that “it is a sterilizing agent and a ploy to control the Muslim population.”
Politicization as exemplified above has unwelcome consequences creating confusion and negativity around a scientific issue (Bolsen & Palm, 2022; Fowler, Nagler, Banka, & Gollust, 2022). A scientific issue is politicized when political interests shape the presentation of scientific facts (Bolsen, Druckman, & Cook, 2014), which was the case demonstrated above. It is known that political actors polarize public attitudes, while nonpartisan experts such as scientists alleviate polarization (Flores et al., 2022). Partisan polarization has emerged as a public health issue in WEIRD nations such as the United States, affecting beliefs and behaviors about COVID-19, health, and related policies (Boxell et al., 2020). Such polarization impedes public support for crisis-tackling policies as well (Flores et al., 2022).
When media coverage becomes politicized and polarized, it has been shown to be associated with intensification of partisan differences in how the public perceives and responds to risks, with individuals aligning their opinions with trusted political elites, even if conflicting with expert information (Druckman, Peterson, & Slothuus, 2013; Garimella, Smith, Weiss, & West, 2021; Modgil, Singh, Gupta, & Dennehy, 2021). Since news media inform, influence, and shape public opinion (Chinn, Hart, & Soroka, 2020) at a time of crisis of massive magnitude (Mach et al., 2021), politicizing and polarizing content from the media leads to overall politicization and polarization of the crisis (Pearman et al., 2021).
We have already discussed why India is an excellent testbed among the non-WEIRD countries to study the phenomena of politicization and polarization in the media. Studies on coverage of news in India have pointed toward differences in framing and slanting (Narayana, & Kapur, 2011), negative coverage of immunization (Das, Singh, & Sharma, 2021), the process of “othering” (Amanullah, Nadaf, & Neyazi, 2023), and biased coverage (Narayana & Kapur, 2011). Since COVID-19 provides us an excellent backdrop to study politicization and polarization of science, we find it imperative to study the extent of politicization and polarization of the COVID-19 crisis through the content analysis of Indian news media during the crisis. Traditional media outlets like newspaper continue to play a central role in disseminating political information, shaping news coverage, and serving public interest, especially during crises like the COVID-19 pandemic, in societies with pluralistic media systems (Mach et al., 2021; Schwoon, Schoeneborn, & Scherer, 2022). Therefore, we have chosen six prominent Indian newspapers to study the extent of politicization and polarization of COVID-19 news coverage.
Ours is the first study to quantitatively measure politicization and polarization of COVID-19-related news coverage based on an extensive dataset of nearly a quarter million articles from six prominent newspapers over a span of 2 years. To our knowledge, no prior study on polarization and politicization of newspaper articles regarding COVID-19 and associated issues of vaccine and lockdown has considered such a large duration or so many articles for any country.
We calculate the mean mention of politicians vis-à-vis scientists in newspaper coverage on COVID-19 as a measure of politicization. We have found a significant level of politicization in the news coverage of COVID-19 vaccine, lockdown, and overall COVID-19 issues. The language difference—the measure of polarization—between the two major national political parties of India, Bharatiya Janata Party (BJP) and Indian National Congress (Congress), on COVID-19 maintains a consistent difference, with an increasing trend during election seasons and a decreasing trend during the peak of COVID-19 waves. On the COVID-19 vaccine issue, the trend is similar, but on the COVID-19 lockdown issue, the degree of polarization has diminished with time. Overall, the politicization and polarization trends in India, resemble a non-uniform pattern, which may be attributed to the perceived shocks in the political competition environment through different state assembly elections.
The rest of our paper is organized as follows. Section on 'Background and Research Question' discusses comprehensively the background and the prior literature, which helps us understand the research gaps and formulate the research questions. Section on 'Data and Methodology' elaborates on the data and methodology for those research questions. Section on 'Results on Politicization and Polarization' narrates our results on the politicization and polarization of COVID-19 news coverage in India. Section on 'Discussion' discusses the importance of our findings on the backdrop of available theoretical frameworks and 'Conclusion' section concludes the paper.
Background and Research Questions
Politicization and Polarization
Politicization can be understood as a three-dimensional process which involves increasing salience (visibility) of political actors, polarization—both intensity and direction-wise—of the issue, and the expansion of actors (range) and audiences involved in the matter (De Wilde, Leupold, & Schmidtke, 2016; Grande, & Hutter, 2016; Kriesi, 2016). Salience may be measured by the number of newspaper articles reporting on it over time. The expansion of actors may be calibrated by examining how frequently these actors have been mentioned in the relevant media coverage.
Polarization—a dimension of politicization—may be defined in terms of both alignment and divergence (Lelkes, 2016; Rekker, 2021). On the alignment account, polarization may be viewed as the popular positioning over an issue defined by their ideological or partisan placement, while on the divergence account, popular positioning happens to oppose an ideological or partisan placement (Rekker, 2021). Polarization may refer to the differing positions taken by various political actors on any issue.
Politicization and Polarization of COVID-19
COVID-19 has posed a challenge to interpret scientific understanding to the public. Science, perceived as an unbiased understanding of reality, becomes politicized when political interests shape the presentation of scientific facts (Bolsen et al., 2014). The politicization of science manifests in various ways, including actor-based politicization, intention-based politicization, and language-based politicization (Bolsen & Palm, 2022; Chinn et al., 2020; Zhou, Liu, & Yang, 2023). This politicization has unwelcome consequences, hindering effective science communication, increasing confusion, and fostering negative responses (Bolsen & Palm, 2022; Fowler et al., 2022). The news coverage about science also increasingly suffers from partisan biases (Hart, Chinn, & Soroka, 2020; Shultziner & Stukalin, 2021). Polarization over science is a multifaceted phenomenon, influenced by psychological differences, denialism, distrust in science, media landscape changes, and partisan preferences and knowledge differences among partisans (Chinn & Pasek, 2021; Hart & Nisbet, 2012; Nisbet, Cooper, & Garrett, 2015; Rekker, 2021).
Partisan polarization has emerged as a public health issue, notably in the United States, affecting beliefs and behaviors about COVID-19, health, and related policies (Boxell et al., 2020). Political polarization impedes public support for crisis-tackling policies (Flores et al., 2022). The COVID-19 pandemic has both decreased and aggravated political division’s impact, with supporters of political ingroups withholding support from the outgroup (Boxell et al., 2020; Stoetzer et al., 2023). Cues from political elites polarize public attitudes toward COVID-19 policies, while cues from nonpartisan experts alleviate polarization (Flores et al., 2022). The politicization and polarization of the COVID-19 vaccine have led to hesitancy and trust issues, influenced by political engagement and endorsement dynamics (Dolman, Fraser, Panagopoulos, Aldrich, & Kim, 2023; Hotez et al., 2021; Xie, Wang, & Ma, 2023).
Regarding COVID-19, the related policy matters of vaccination and lockdown have also manifested politicization and polarization. While developing a vaccine is pivotal, partisanship has been showed to hinder pandemic mitigation (Ward et al., 2020). Vaccinated individuals may show polarized attitudes toward the unvaccinated, leading to intergroup conflict (Henkel, Sprengholz, Korn, Betsch, & Böhm, 2023). The politicization of vaccine approval decreases confidence, while expert endorsement increases it (Bokemper, Huber, Gerber, James, & Omer, 2021). Lockdown execution during COVID-19 remained contentious among politicians, citizens, and media coverage (Li, Erfani, Wang, & Cui, 2021; Zhang, 2021).
Media’s Role in Shaping Public Opinion, Especially During COVID-19
Evidence rules out any suspicion that the traditional media does not play an important role in the age of digital and social media (Schwoon et al., 2022). Nielsen, Fletcher, Newman, Brennen, and Howard (2020) observed that people perceive news media as instrumental in their understanding of the pandemic, with mainstream news media being a particularly trusted source of COVID-related information (Mont’Alverne et al., 2022). Studies show that choosing a media piece itself is an act affected by existing polarization, and such media contributes to exacerbate polarization in turn (Garimella et al., 2021; Modgil et al., 2021). Ejaz, Ittefaq, Seo, and Naz (2021) show how trust in media contributed to the conspiracy theories developed during the pandemic. During the pandemic, the media’s portrayal of the dynamics of epidemiological science has remarkably influenced public understanding of health risks, health policy measures, and associated political and policy discussions (Mach et al., 2021) as the news media is the primary channel for the public to know about the issue (Kasperson et al., 1988). News reports on issues like face mask wearing as a safety measure have witnessed politicization and polarization, breeding doubts and eroding trust in public health authorities (Lang, Erickson, & Jing-Schmidt, 2021; Pascual-Ferrá et al., 2021).
Inadequate scientific quality in pandemic news coverage has historically impeded public health efforts (Hoffman & Justicz, 2016). Biases in newsroom norms and partisanship in media (Borah, Ghosh, Hwang, Shah, & Brauer, 2023); the desire to draw audience attention (Boykoff, 2011); personalized stories focusing on arguments between competing actors to highlight conflict; perpetuating narratives and dramatizing issues (Feldman et al., 2017; Hart & Feldman, 2014); prominently featuring political actors as official sources rather than scientists (Bolsen et al., 2014), all these lead to politicization and polarization of content and their partisan framing. Politicized and polarized coverage can influence public views by encouraging individuals to follow political elites’ opinions rather than those of scientists (Bolsen et al., 2014). Motivated reasoning, as identified by Taber et al. (2009), coupled with the heightened influence of partisan elites on individuals’ attitudes toward issues, as highlighted by Druckman et al. (2013), are likely to result in the formation of public opinions that align with the political elites they trust. This tendency leads individuals to dismiss information that does not conform to their established views. In summary, media coverage plays a crucial role in shaping public opinion, particularly in the context of emerging science and risk such as COVID-19.
A study by Mihelj et al. (2022) contends that the COVID-19 pandemic constituted a disruptive, unpredictable, and exhausting media event with a high potential for divisiveness and polarization, particularly in cases of elite-led politicization, low media freedom, and declining democratic standards. During the first wave of the crisis, there was a notable increase in news consumption among the public (Mihelj et al., 2022). Hart et al. (2020) have identified evidence of pandemic polarization, even in the early stages, as reported by mass media in the United States. Wichowsky and Condon (2022) have found in their U.S. based study that media coverage highlighting partisan divides on COVID-19 reinforced polarization, while emphasizing common concerns across party lines led to less polarized responses. Palau-Sampio (2021) has highlighted the emergence of an “infodemic”—a global epidemic of disinformation—due to COVID-19, underscoring the links between political ideology, polarized discourse, populism, and the spread of misinformation.
India in the Context of Politicization and Polarization of COVID-19
During the COVID-19 pandemic, India has witnessed a dramatic presentation of the crisis with politically charged and catchy phrases, creating a climate of opinion (Biswas et al., 2021) with media reporting on divisive lines (Quarcoo & Kleinfeld, 2020, 2022) through the process of “othering” (Amanullah et al., 2023) and biased coverage (Sharma & Anand, 2020). All these may have led to spiraling effect of ostracization (Biswas et al., 2021) and affected the larger discourse about community (Amanullah et al., 2023).
In the aftermath of the pandemic, India has experienced a significant degree of polarization and politicization associated with socio-economic issues (Akbar, Panda, Kukreti, Meena, & Pal, 2021). There is an evident affective polarization based on social identity, although there is a notable absence of strong political party identification (Arabaghatta Basavaraj et al., 2021). Additionally, there has been a trend of politicization and vilification of communities in relation to the COVID-19 issue (Biswas, Chatterjee, & Sultana, 2021). Overall, India may have experienced a heightened level of political polarization during this pandemic (Sahoo, 2020), which makes it important to understand the role of media toward such politicization and polarization.
Borah and Singh (2022) and Dash, Mishra, Shekhawat, and Pal (2022) have analyzed content from the social media platform Twitter to have found the COVID-19 issue as less polarized compared to others. A few studies pertaining to the COVID-19 vaccine, such as Paliwal, Parveen, Alam, and Ahmed (2022) and Verma, Chhabra, and Gupta (2022) have found using the Twitter data, changing public opinion over time but have not precisely dealt with politicization or polarization of the underlying issue(s) that might have driven such change in public opinion.
In this context, we have identified the following research gap. While some studies have looked into different issues related to the COVID-19 response in India; however, very few studies have considered COVID-19 news coverage in a quantitative framework. Moreover, no study, to our knowledge, has measured the issue of politicization and polarization based on COVID-19 news coverage from the newspapers for a comprehensive 2-year long period.
Research Questions
Our paper seeks to enquire into the extent to which mainstream news outlets represent political actors against scientific opinions in the context of COVID-19 coverage. We also examine the degree to which language in news coverage accentuates partisan differences in the Indian context.
This study addresses two overarching questions:
RQ1: Does the news coverage of COVID-19 in mainstream news media in India exhibit politicization?
In pursuing an answer to this primary question, the study further delves into the following sub-questions:
RQ1a: To what extent is COVID-19 news coverage politicized in India across time periods?
RQ1b: To what extent is news coverage on vaccines for COVID-19 politicized in India across time periods?
RQ1c: To what extent is news coverage on lockdown due to COVID-19 politicized in India across time periods?
Polarization is the most interesting component of politicization (De Wilde et al., 2016). While the politicization trends offer valuable information on the extent of politicians being evoked in COVID-19 news coverage, polarization trends offer insights into the distinctions among politicians from different political parties in terms of their language difference and its dynamics.
RQ2: Does news coverage of COVID-19 in mainstream news media in India exhibit polarization while discussing political parties?
The study extends its investigation to address the following related questions:
RQ2a: What is the extent of partisan differences in the language of news coverage of COVID-19 in India when discussing political parties across time periods?
RQ2b: What is the extent of partisan differences in the language of news coverage of vaccines of COVID-19 in India when discussing political parties across time periods?
RQ2c: What is the extent of partisan differences in the language of news coverage of lockdown for COVID-19 in India when discussing political parties across time periods?
In sum, the primary objective of our study is to discern the trends in the politicization and polarization of COVID-19 news coverage in India. Describing the trends in politicization and polarization in COVID-19 news coverage is imperative in elucidating the factors contributing to the polarization of public opinion in India regarding the crisis and the fatal health risks posed by COVID-19.
Data and Methodology
Description of the Dataset
Our dataset comprises news coverage from six prominent Indian newspapers—The Times of India, Hindustan Times, The Hindu, Indian Express, The Telegraph, and India Today Online—collected from Lexis-Nexis during January 2020 to April 2022. We have chosen this extended time period to capture the entirety of the COVID-19 pandemic and the three COVID-19 waves India faced (see Supplementary Appendix 1), with the third wave persisting till March and also have accommodated the idea of fatigue in media coverage of COVID-19 (Pearman et al., 2021).
The reason for choosing a duration of more than 2 years lies in the changing situation over time and over different waves (Pachauri & Pachauri, 2023). Since newspapers frame news items differently (Mutua & Oloo, 2020; Ogbodo et al., 2020), with a degree of slant varying from newspaper to newspaper (Narayana & Kapur, 2011) relying upon just one or two newspapers could, arguably, lead to a biased analysis. In the interest of a more balanced analysis, we have taken six prominent newspapers in India. All these lead to our complete dataset for COVID-19-related news having 231,519 stories, identified through the mention of terms like Corona, Coronavirus, or Covid (Table 1).
Number of News Articles for Analysis
| Dictionary | Number of articles |
|---|---|
| COVID-19 | 2,31,519 |
| Covid-Political actors (combined) | 67,028 |
| Covid-Scientists | 72,714 |
| Covid-Vaccine | 41,424 |
| Vaccine-Political actors (combined) | 15,256 |
| Vaccine-Scientists | 19,514 |
| Covid-Lockdown | 35,021 |
| Lockdown-Political actors(combined) | 11,378 |
| Lockdown-Scientists | 5102 |
| Dictionary | Number of articles |
|---|---|
| COVID-19 | 2,31,519 |
| Covid-Political actors (combined) | 67,028 |
| Covid-Scientists | 72,714 |
| Covid-Vaccine | 41,424 |
| Vaccine-Political actors (combined) | 15,256 |
| Vaccine-Scientists | 19,514 |
| Covid-Lockdown | 35,021 |
| Lockdown-Political actors(combined) | 11,378 |
| Lockdown-Scientists | 5102 |
Note. There is some overlapping of articles among the groups of politicians and scientists, as in some articles, both of them are mentioned.
Number of News Articles for Analysis
| Dictionary | Number of articles |
|---|---|
| COVID-19 | 2,31,519 |
| Covid-Political actors (combined) | 67,028 |
| Covid-Scientists | 72,714 |
| Covid-Vaccine | 41,424 |
| Vaccine-Political actors (combined) | 15,256 |
| Vaccine-Scientists | 19,514 |
| Covid-Lockdown | 35,021 |
| Lockdown-Political actors(combined) | 11,378 |
| Lockdown-Scientists | 5102 |
| Dictionary | Number of articles |
|---|---|
| COVID-19 | 2,31,519 |
| Covid-Political actors (combined) | 67,028 |
| Covid-Scientists | 72,714 |
| Covid-Vaccine | 41,424 |
| Vaccine-Political actors (combined) | 15,256 |
| Vaccine-Scientists | 19,514 |
| Covid-Lockdown | 35,021 |
| Lockdown-Political actors(combined) | 11,378 |
| Lockdown-Scientists | 5102 |
Note. There is some overlapping of articles among the groups of politicians and scientists, as in some articles, both of them are mentioned.
To ensure substantive coverage of COVID-19, we have restricted our analysis to articles with at least three mentions of the COVID-19 keywords. For articles related to COVID-19 lockdown, we have identified them through the occurrence of lockdown keywords within sentences mentioning COVID-19 keywords. Similarly, for articles related to the COVID-19 vaccine, we have identified them through the occurrence of vaccine keywords within sentences mentioning COVID-19 keywords. Since manual analysis of these many news articles is difficult, costly, and ridden with the problem of intense supervision, we have chosen the more efficient method of computer-based content analysis (Chinn et al., 2020; De Graaf & van der Vossen, 2013; Hart et al., 2020; Zhou et al., 2023), using R and Python for computational purposes.
Measuring Politicization
In our framework, the presence of political actors in a related news article is central to the conceptualization of politicization defined through salience and expansion of actors in the literature. We measure the relative frequency of political actors (alternatively, called politicians) over time in the news articles related to the COVID-19 pandemic as the measure of politicization. More explicitly, we have prepared a dictionary of proper and common nouns representing political actors and non-political actors such as scientists (see Appendix A) and employed a dictionary-based content analysis method. We categorized political actors into three major party blocks: the first representing the ruling party, “Bhartiya Janata Party” (henceforth, BJP); the second representing the principal opposition party, “Indian National Congress” (henceforth, Congress); and the third representing multiple national and regional parties playing important roles in national and regional politics, referred to as “Other parties” (Appendix Table A-1). While “Other parties” include various political parties with different political orientations, we grouped them together for practical purposes.
Dictionaries for Analysis
| Dictionary | Words |
|---|---|
| Covid-19 | Corona, Coronavirus, Covid |
| Scientist | Scientist, Research, Professor, Doctor, Health official, Health service, Dr, Health authority, World Health Organization, CDC, Indian medical association, AIIMS, Sero survey, Health Secretary, union Health secretary, ICMR, CDSCO, NCDC, Randeep Guleria |
| BJP | BJP, Rashtriya Swayamsevak Sangh, RSS, Modi, Amit Shah, Yogi, Harsh Vardhan, Mandaviya, Bhartiya Janata Party, VHP, Bajrang Dal, BJYM, Nirmala Sitharaman, Himanta, Nadda, Sambit Patra, Rajnath Singh, Jaishankar, Sudhanshu Trivedi |
| Congress | Rahul Gandhi, Sonia Gandhi, Priyanka Gandhi, Congress leader, Congress Spokesperson, Congress Party, Shashi Tharoor, Congress, Indian National Congress, NSUI, AICC, Adhir Ranjan Chowdhury, Mallikarjun Kharge, Randeep Surjewala, Bhupesh Baghel, Kamal Nath, Gehlot, Manmohan Singh, Siddaramaiah |
| Other Parties | KK Shailaja, Shiv Sena, BSP, NCP, Sharad Pawar, YSR Congress, Sitaram Yechuri, AIDMK, DMK, CPIM, AITC, Trinmool, AAP, Manish Sisodiya, Pinarayi Vijayan, M. K. Stalin, Mamata Banerjee, Uddhav Thackeray, Arvind Kejriwal, Naveen Patnaik, K Chandrashekar Rao, YS Jaganmohan Reddy, Aam Aadmi Party, Samajwadi Party, RJD, JDU, Nitish Kumar, BJD, Akhilesh Yadav, Communist Party of India |
| Government | Central government, Government of India, State government |
| Vaccine | Vaccine, Covishield, Covaxin, Sputnik, SII, Serum Institute of India, Cowin, Astrazeneca, Moderna, Pfizer, Oxford Astrazeneca, Bharat Biotech, Adar Poonawala |
| Lockdown | Safer at home, stay at home, lockdown, lock-down, travel ban |
| Other terms for BJP | Ruling party, Right-wing, Nationalist, Hindu-nationalist |
| Other terms for Congress | Opposition party, centrist, secular, grand old party |
| Other parties | Left-wing, Dravidian party |
| Dictionary | Words |
|---|---|
| Covid-19 | Corona, Coronavirus, Covid |
| Scientist | Scientist, Research, Professor, Doctor, Health official, Health service, Dr, Health authority, World Health Organization, CDC, Indian medical association, AIIMS, Sero survey, Health Secretary, union Health secretary, ICMR, CDSCO, NCDC, Randeep Guleria |
| BJP | BJP, Rashtriya Swayamsevak Sangh, RSS, Modi, Amit Shah, Yogi, Harsh Vardhan, Mandaviya, Bhartiya Janata Party, VHP, Bajrang Dal, BJYM, Nirmala Sitharaman, Himanta, Nadda, Sambit Patra, Rajnath Singh, Jaishankar, Sudhanshu Trivedi |
| Congress | Rahul Gandhi, Sonia Gandhi, Priyanka Gandhi, Congress leader, Congress Spokesperson, Congress Party, Shashi Tharoor, Congress, Indian National Congress, NSUI, AICC, Adhir Ranjan Chowdhury, Mallikarjun Kharge, Randeep Surjewala, Bhupesh Baghel, Kamal Nath, Gehlot, Manmohan Singh, Siddaramaiah |
| Other Parties | KK Shailaja, Shiv Sena, BSP, NCP, Sharad Pawar, YSR Congress, Sitaram Yechuri, AIDMK, DMK, CPIM, AITC, Trinmool, AAP, Manish Sisodiya, Pinarayi Vijayan, M. K. Stalin, Mamata Banerjee, Uddhav Thackeray, Arvind Kejriwal, Naveen Patnaik, K Chandrashekar Rao, YS Jaganmohan Reddy, Aam Aadmi Party, Samajwadi Party, RJD, JDU, Nitish Kumar, BJD, Akhilesh Yadav, Communist Party of India |
| Government | Central government, Government of India, State government |
| Vaccine | Vaccine, Covishield, Covaxin, Sputnik, SII, Serum Institute of India, Cowin, Astrazeneca, Moderna, Pfizer, Oxford Astrazeneca, Bharat Biotech, Adar Poonawala |
| Lockdown | Safer at home, stay at home, lockdown, lock-down, travel ban |
| Other terms for BJP | Ruling party, Right-wing, Nationalist, Hindu-nationalist |
| Other terms for Congress | Opposition party, centrist, secular, grand old party |
| Other parties | Left-wing, Dravidian party |
Dictionaries for Analysis
| Dictionary | Words |
|---|---|
| Covid-19 | Corona, Coronavirus, Covid |
| Scientist | Scientist, Research, Professor, Doctor, Health official, Health service, Dr, Health authority, World Health Organization, CDC, Indian medical association, AIIMS, Sero survey, Health Secretary, union Health secretary, ICMR, CDSCO, NCDC, Randeep Guleria |
| BJP | BJP, Rashtriya Swayamsevak Sangh, RSS, Modi, Amit Shah, Yogi, Harsh Vardhan, Mandaviya, Bhartiya Janata Party, VHP, Bajrang Dal, BJYM, Nirmala Sitharaman, Himanta, Nadda, Sambit Patra, Rajnath Singh, Jaishankar, Sudhanshu Trivedi |
| Congress | Rahul Gandhi, Sonia Gandhi, Priyanka Gandhi, Congress leader, Congress Spokesperson, Congress Party, Shashi Tharoor, Congress, Indian National Congress, NSUI, AICC, Adhir Ranjan Chowdhury, Mallikarjun Kharge, Randeep Surjewala, Bhupesh Baghel, Kamal Nath, Gehlot, Manmohan Singh, Siddaramaiah |
| Other Parties | KK Shailaja, Shiv Sena, BSP, NCP, Sharad Pawar, YSR Congress, Sitaram Yechuri, AIDMK, DMK, CPIM, AITC, Trinmool, AAP, Manish Sisodiya, Pinarayi Vijayan, M. K. Stalin, Mamata Banerjee, Uddhav Thackeray, Arvind Kejriwal, Naveen Patnaik, K Chandrashekar Rao, YS Jaganmohan Reddy, Aam Aadmi Party, Samajwadi Party, RJD, JDU, Nitish Kumar, BJD, Akhilesh Yadav, Communist Party of India |
| Government | Central government, Government of India, State government |
| Vaccine | Vaccine, Covishield, Covaxin, Sputnik, SII, Serum Institute of India, Cowin, Astrazeneca, Moderna, Pfizer, Oxford Astrazeneca, Bharat Biotech, Adar Poonawala |
| Lockdown | Safer at home, stay at home, lockdown, lock-down, travel ban |
| Other terms for BJP | Ruling party, Right-wing, Nationalist, Hindu-nationalist |
| Other terms for Congress | Opposition party, centrist, secular, grand old party |
| Other parties | Left-wing, Dravidian party |
| Dictionary | Words |
|---|---|
| Covid-19 | Corona, Coronavirus, Covid |
| Scientist | Scientist, Research, Professor, Doctor, Health official, Health service, Dr, Health authority, World Health Organization, CDC, Indian medical association, AIIMS, Sero survey, Health Secretary, union Health secretary, ICMR, CDSCO, NCDC, Randeep Guleria |
| BJP | BJP, Rashtriya Swayamsevak Sangh, RSS, Modi, Amit Shah, Yogi, Harsh Vardhan, Mandaviya, Bhartiya Janata Party, VHP, Bajrang Dal, BJYM, Nirmala Sitharaman, Himanta, Nadda, Sambit Patra, Rajnath Singh, Jaishankar, Sudhanshu Trivedi |
| Congress | Rahul Gandhi, Sonia Gandhi, Priyanka Gandhi, Congress leader, Congress Spokesperson, Congress Party, Shashi Tharoor, Congress, Indian National Congress, NSUI, AICC, Adhir Ranjan Chowdhury, Mallikarjun Kharge, Randeep Surjewala, Bhupesh Baghel, Kamal Nath, Gehlot, Manmohan Singh, Siddaramaiah |
| Other Parties | KK Shailaja, Shiv Sena, BSP, NCP, Sharad Pawar, YSR Congress, Sitaram Yechuri, AIDMK, DMK, CPIM, AITC, Trinmool, AAP, Manish Sisodiya, Pinarayi Vijayan, M. K. Stalin, Mamata Banerjee, Uddhav Thackeray, Arvind Kejriwal, Naveen Patnaik, K Chandrashekar Rao, YS Jaganmohan Reddy, Aam Aadmi Party, Samajwadi Party, RJD, JDU, Nitish Kumar, BJD, Akhilesh Yadav, Communist Party of India |
| Government | Central government, Government of India, State government |
| Vaccine | Vaccine, Covishield, Covaxin, Sputnik, SII, Serum Institute of India, Cowin, Astrazeneca, Moderna, Pfizer, Oxford Astrazeneca, Bharat Biotech, Adar Poonawala |
| Lockdown | Safer at home, stay at home, lockdown, lock-down, travel ban |
| Other terms for BJP | Ruling party, Right-wing, Nationalist, Hindu-nationalist |
| Other terms for Congress | Opposition party, centrist, secular, grand old party |
| Other parties | Left-wing, Dravidian party |
We have calculated the mean number of mentions of political actors in the news articles over time. This statistic allows for the comparison of politicization across issues over time and can be applied to a large dataset, as demonstrated by Chinn et al. (2020) and Hart et al. (2020). To understand the mention of political actors in COVID-19 news coverage in the given context, we have also calculated the frequency of mentions of scientists in news articles. The idea is to examine the relative emphasis of the political perspective vis-à-vis the scientific perspective through the mean number of mentions of political actors and scientists, as COVID-19 has involved both public and scientific risk factors.
To comprehensively measure the politicization of COVID-19 coverage, we need to examine some crucial issues like the COVID-19 vaccine and the COVID-19 lockdown. We have examined the differential nature of politicization regarding these crucial issues as opposed to the general politicization around COVID-19. We have followed the same dictionary for political actors and scientists and the same statistics to measure the underlying politicization.
Measuring Polarization
To measure the trend of polarization in the COVID-19 news over time, we appeal to the Wordfish—a Poisson scaling model—commonly used to estimate polarization (Slapin & Proksch, 2008). The Wordfish model assigns each text document a score based on the ideological position—also described as a score for language difference by Chinn et al. (2020)—of the text. This model effectively assumes a uni-dimensionality principle under which the latent principal dimension extracted from a text document represents the political content of the underlying document (Goet, 2019). In other words, the estimated statistic under this model is the ideological position of the document in that latent principal dimension extracted.
It is important to note that this Wordfish method requires neither prior information about the positions of the actors nor requires reference texts. Therefore, the Wordfish estimates of polarization for different text documents over time are only dependent upon the content itself and independent of the timing of the document. Our purpose of comparison over time requires a uniform benchmarking for measuring polarization over time, which is offered by the Wordfish method.
We appeal to the average value of the Wordfish statistic representing the ideological position of the news articles related to a political actor. The difference in ideological position, as expressed through the difference of these average values for two opposing groups of political actors—henceforth referred to as language difference—indicates the extent of polarization on the issue. Understandably, polarization increases if this language difference increases over time.
As BJP and Congress both represent the two poles of the political spectrum (Chris, 2012; Zavos, 2000). We have not included the other political parties in the primary analysis of this paper as there are a number of national and state-level parties in India who, together, do not represent any coherent ideological bloc. However, for interested readers, we have included the relevant results in Supplementary Appendix 3. We have focused our analysis based on their language difference in the news coverage as the measure of polarization as one coherent number out of our statistical model. We have identified COVID-19 news articles that mention BJP but no other parties. We have, then, removed all mentions of BJP from those articles so that that word should not factor in the assessment of the language difference. We have similarly processed other sets of articles mentioning only Congress but no other parties. We have calculated the Wordfish statistic for language difference for each of the processed articles. The average value of the statistic for the particular set of articles defines the party position.
Results on Politicization and Polarization
Politicization of COVID-19 Vaccine News Coverage
For the entire period of analysis for COVID-19 vaccine news—March 2020 to April 2022—the mean mention of politicians per document at 1.34 has remained lower than that of scientists at 1.40, indicating more salience given to scientists compared to politicians. However, the expansion of politicians has happened over time, implying increasing politicization of the COVID-19 vaccine. The mean mention of politicians is higher at 1.37 compared to that of scientists at 1.14 during March 2021–April 2022. The larger mention of politicians is connotative of the politicization of the issue. Interestingly, this is exactly the time period of the rollout of the vaccines in India to the general public in phases, indicating the politicization of vaccine, right after being rolled out.
As far as the statistical significance of this difference is concerned, we have performed a one-tailed Student’s t-test (t-statistic: 4.83; p-value < .01) for comparing two populations with different variances for this time period of March 2021–April 2022. As we have found the mean mention of politicians significantly larger than that of the scientists at 1% significance level, we conclude an extremely significant level of politicization.
During October 2020–November 2020, the mean mention of politicians reaches a sudden peak (Figure 1). This period coincides with the state assembly election for the Indian state of Bihar (henceforth, Bihar assembly election) in which COVID-19 vaccine has figured in the manifestos of political parties (Hebbar, 2020). Again during April 2021–June 2021, the mean mention of politicians attains another peak at the time of the most fatal second wave of COVID-19 in India. The time periods of a jump in politicization seem to coincide with politically important moments.
Politicization in newspaper coverage of COVID-19 Vaccine from March 2020 to April 2022.
Politicization of COVID-19 Lockdown News Coverage
For the entire period of our analysis of the COVID-19 lockdown during March 2020 to March 2022, the mean mention of politicians remains remarkably (t-statistic: 7.13; p-value < .01) higher at 1.23 compared to that of scientists at 0.36. Interestingly, politicians’ mean mention has decreased significantly during June 2021–November 2021 (Figure 2).
Politicization in Newspaper Coverage of COVID-19 lockdown from March 2020 to March 2022.
Politicization of Overall COVID-19 News Coverage
For our entire period—from January 2020 to March 2022—the mean mention of politicians (political actors) per document is 0.97 compared to the corresponding number of 0.95 for scientists (non-political actors). These figures indicate a greater salience for the politicians compared to the scientists on the COVID-19 issue. We have plotted the mean mentions for politicians and scientists alike for each week as well as 4-week moving averages in Figure 3. We do not observe any secular trend over time as far as the expanse of politicization is concerned.
Politicization in newspaper coverage of COVID-19 during January 2020 to March 2022.
For the first year—January 2020 to December 2020—the mean mention of scientists at 0.96 is greater than that of politicians at 0.91. However, during January 2021—March 2022, a period of two waves of COVID-19, the mean mention of politicians has become higher at 1.05 compared to that of scientists at 0.95. During the initial months, the mean mention of politicians is less than the corresponding figure for scientists, but by the beginning of April 2020, the mean mention of politicians has become larger than that of scientists.
While there are ups and downs regarding mentions of politicians and scientists in the news coverage of COVID-19, we can divide the entire duration into sub-periods: Pre-launch of COVID-19 vaccine for all and post-launch of COVID-19 vaccine for all. For the initial period of January 2020—May 2022, the mean mention of politicians was 0.97 against the corresponding figure of 0.94 for scientists. This difference was found to be statistically significant in the t-test (t-statistic: 2.02; p-value < .05).
Polarization in COVID-19 Vaccine News Coverage
The degree of polarization between BJP and Congress on the issue of COVID-19 vaccine has gone through huge fluctuations during the initial months (Figure 4). The degree of polarization has intensified during October 2020–coinciding with the Bihar assembly election—but then takes a sudden dip in November 2020. We can see a U-shaped trend in the language difference between BJP and Congress during February 2021 to August 2021. Interestingly, the vaccination started in India in January 2021 for the health and frontline workers and was subsequently rolled out for the general public in phases since March 2021. During the height of the second wave of COVID-19 in India, polarization had come down but again increased after the end of the second wave.
Polarization in COVID-19 vaccine news coverage from March 2020 to April 2022.
Polarization in COVID-19 Lockdown News Coverage:
In the very first month of the announcement of the nationwide lockdown in March 2020, the language difference between BJP and Congress is found to be high, but then onwards, it has started declining. By July 2020, the difference reached a low that is, statistically speaking, no different than zero (Figure 5). Again, this difference has risen to its highest peak in October 2020, coinciding with the Bihar assembly election.
Polarization in COVID-19 lockdown news coverage from March 2020 to July 2021.
Polarization in Overall COVID-19 News Coverage
On overall COVID-19 coverage, we have plotted the language difference between BJP and Congress in Figure 6. For the entire period of our analysis—during March 2020 to April 2022—we found that the degree of polarization varies over time, particularly increasing during the time period during election season—from one or two months before the beginning of a state election (around the time of notification of elections) to at least till its completion. The language difference between BJP and Congress remained stable during the initial period of the pandemic but peaked in September 2020 on the eve of the Bihar assembly election. This language difference decreases only after the effective completion of the state election in five different states Assam, Kerala, Puducherry, Tamil Nadu, and West Bengal in April 2021.
Polarization in COVID-19 news coverage from March 2020 to April 2022.
Long-Term Trends in Polarization
Political parties can change their stance on a particular issue for a comparatively shorter time period, but in a longer time period, they, arguably, remain rooted to their political ideology. The analysis of differences in language for a longer time period may reflect this difference in political ideology between different political groups. We have, therefore, done a comparatively long-term analysis based on 3-month and 6-month data. The reader can refer to Supplementary Appendix 4 for a detailed analysis. We highlight the summary of our analysis based on 6-month data here (Figure 7). Our analysis based on the three-month data has shown common features found both in the analysis of the monthly data and the 6-monthly data.
Long-term polarization in COVID-19, Covid-Vaccine and lockdown news coverage.
COVID-19 news coverage-wise, the extent of polarization between BJP and Congress remains comparatively lower. However, the language difference between BJP and Congress remains significant and persistent over the entire horizon of our analysis. The very overall pattern is observed on the issue of COVID-19 lockdown, with the language difference between BJP and Congress remaining even lower in magnitude. On the COVID-19 vaccine issue, the language difference between BJP and Congress has been fluctuating but generally at a higher level of difference compared to the earlier issues. We observed an increase in Polarization between BJP and Congress during the period of March 2020 to Feb 2021, but during the period of March–August 2021, we observed a dip in polarization; this period witnessed the deadly second wave of Covid in India and also the roll-out of the vaccine program. We can observe both prominent parties coming closer ideologically on vaccine during this period.
Discussion
How public opinion is influenced by the polarization among the political parties can depend on the extent to which elite communication encourages social categorization. In our paper, we have investigated and analyzed the level of politicization and polarization of science within the source of such elite communications, namely newspaper reports mentioning politicians or scientists, both of whom, are, arguably, social elites. More crucially, political responses to a crisis like COVID-19 and the public response to those political cues decide the way a nation will move forward in tackling the crisis.
Fundamentally, there are two ways for a nation to move toward. If opinions in political discourse on how to handle the COVID-19 crisis continue to diverge, this could very well worsen affective polarization on a micro level (Green, Edgerton, Naftel, Shoub, & Cranmer, 2020). Conversely, there can emerge a unifying force or language from all political actors to bring people together against a common enemy (Grossman, Kim, Rexer, & Thirumurthy, 2020; Lenz, 2012, Merkley & Stecula, 2018). On the issue of COVID-19, we have found—in India—a curious mix of both instances. The polarization has—remarkably—augmented during election seasons. On the other hand, the polarization has decreased during the time of second COVID-19 wave which was the most difficult time for India during the pandemic.
During October 2020–November 2020, when the political blocs encountered Bihar assembly election—the first prominent political competition during COVID-19 pandemic—polarization of COVID-19 and related issues registered a prominent upward jump. At the same time, politicization of those issues has witnessed distinct jumps, sometimes even more sharply than polarization. Interestingly, during March 2021–May 2021 when the assembly elections in five Indian states partially coincided with a larger period of deadly COVID-19 second wave, we have observed mostly a declining trend in polarization among the political blocs on the issue of COVID-19 and an increasing trend in the politicization of the related issues. This is paradoxical given the fact that polarization is considered a component of politicization (De Wilde et al., 2016; Kriesi, 2016). We found that although during the biggest crisis period of the second wave, salience—as measured by mean mentions—has increased, but the language difference among the political blocs has become less prominent. Increasing politicization, in a time of great crisis, may not necessarily imply an augment in polarization for India, as we observe here.
The non-uniform trend in politicization in India—for COVID-19 issues—may be partially attributed to different policy styles—anticipatory or reactive—of the political actors (Blauberger, Heindlmaier, Hofmarcher, Assmus, & Mitter, 2023). Anticipatory policy style was operative in the beginning of the pandemic with public policy attempting to get solutions from the experts to the perceived problem. This style of policy, possibly, had led to lesser politicization during the initial months in India, as we have noted. Reactive policy style responds to the pressure of the problem rather than solving it and lacks broad public consensus—all of which might lead to greater politicization, which is exactly what we have observed in our analysis. Essentially, the politicization of the COVID-19 issues observed by us is non-monotone and varying over time. Scholars have detected this fluctuating pattern of politicization over time in other contexts such as “differentiated politicization” by De Wilde et al. (2016), “intermittent politicization” by Kriesi (2016), and “punctuated politicization” by Grande and Kriesi (2016). Although for the COVID-19 issue, the Indian story resembles both “differentiated politicization” and “intermittent politicization,” for the COVID-19 vaccine issue, it is a case of “punctuated politicization.” In the former scenario, there is no single break point identifiable in the entirety of the time series, whereas for the latter scenario, twin prominent break points may be identified.
To explain the dichotomous nature of polarization in India, we resort to the Spatial model of political competition (Stokes, 1963). Political parties adjust their position on a particular policy with respect to their opposition party and also with respect to the parties of their “ideological families” in a multiparty system (Adams & Somer-Topcu, 2009), valence image of party elites (Adams, 2012; Adams, Clark, Ezrow, & Glasgow, 2004), and in response to shifts in public opinion (Adams et al., 2004). Parties may respond to a situation when a shift in public opinion has become clearly disadvantageous to them but—being reluctant to alter their ideologies—may not respond as well (Adams et al., 2004; Budge, 1994). Our observation of the short-term (monthly) fluctuations in polarization may happen because of shifts in public opinion in the wake of new challenges and new experiences from COVID-19. However, in the long run (six-monthly), political parties may be reluctant to alter their ideology, and therefore, the extent of polarization remains consistent between political blocs.
Political processes are shaped by “friction” between parties, ideologies, and policy choices. Political agendas, consequently, do not adapt to the real-world impulses smoothly, results in either ignoring them or over-reacting to them. On account of this friction, a brief period of intensified partisanship may account for most of the increase in polarization and also can have lasting & significant effects. This pattern in polarization marked by high jumps is called “Punctuated polarization” happening due to real or perceived shock in the political competition environment (Bonica, 2014; Walgrave & Nuytemans, 2009). The peaks in the measure of polarization in our study are often registered during the election seasons indicating real or perceived shock in the environment of political competition. This observation makes the Indian phenomenon of polarization concerning the COVID-19 issues a fit case for “Punctuated polarization.”
Conclusion
Our analysis is based on a very large dataset comprising of news coverage of COVID-19 over 2 years in India. Our analysis concludes massive politicization regarding COVID-19 vaccine and COVID-19 lockdown. The overall COVID-19 coverage is significantly politicized before the roll-out of vaccine for all, which coincided with the end of the second wave of COVID-19. COVID-19 news coverage was significantly polarized between the two prime political blocs of BJP and Congress. Politicization and polarization of science in the Global South seem to be motivated by competition in the political environment as is the case with the democracies in Global North. Interestingly, in the Indian scenario, an increasing politicization in the time of great crisis need not be accompanied by sharpening polarization.
Our study does not explore how non-official political actors, such as political activists, may politicize COVID-19 news by pushing certain public agendas. Additionally, it does not capture the impact of journalists on the politicization of news coverage. Furthermore, we have not considered newspapers published in the Indian languages in our data source, which have more readership compared to the English-language newspapers. We have not covered the electronic media coverage in any language as well. Future studies could develop computationally more efficient methodologies to deal with these challenges.
By investigating instances of politicization and polarization of scientific news, researchers can identify vulnerabilities in the dissemination process and assess the factors that drive the manipulation of information. These insights can inform the development of targeted interventions to improve science communication, bridge the gap between scientific experts and the public, and counteract the divisive forces that threaten the integrity of knowledge dissemination. Furthermore, a nuanced understanding of this phenomenon can contribute to the formulation of evidence-based policies that are less susceptible to political expediency and more attuned to the complex realities of global challenges.
Biographical Note
Swarn Rajan is a PhD scholar pursuing research at Indian Institute of Management Kozhikode. Currently in the final year of the doctoral program, his research interest lies in the realm of public policy, data science and political economy.
Kausik Gangopadhyay is an economist who earned his PhD from the University of Rochester (2007). He is a Professor in the Indian Institute of Management Kozhikode. He has published more than 20 articles in reputed refereed international journals with more than 425 citations in google scholar. He also has a coedited book published by Springer Verlag. He also enjoys writing popular article. He lives in Kozhikode with his wife and two children.
Anirban Ghatak is a faculty of Economics at Indian Institute of Management Kozhikode. Anirban’s research interest lies in the intersection of data science, economics, and human behavior. He has published several articles in reputed international peer reviewed journals and is active in the consulting space for budding entrepreneurs. Anirban loves photography, animals, and teaching.
Funding
The authors received no specific funding for this work.
Conflict of Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A
Dictionaries for Analysis
The development of our dictionary is underpinned by the foundational work of Hart et al. (2020) serving as a primary source of inspiration. Drawing from their methodology, we meticulously adapted their framework to suit the nuances of the Indian political landscape. In the classification of common nouns such as BJP and Congress, our selection of words aligns with prevalent usage patterns in describing these political entities. Furthermore, in the case of proper nouns, encompassing figures such as Narendra Modi, Rahul Gandhi, and Randeep Guleria, we employed a strategy that involves categorizing these leaders based on their official roles, including cabinet ministers, party spokespersons, the face of the party, chief ministers, and directors of esteemed institutions such as AIIMS. Drawing inspiration from Hart et al. (2020), we focus on named political actors considering the limited time period for our analysis. Named political actors are less likely to change during this period and exhibit more consistency in appearing in the coverage, unlike the large time period covered by Chinn et al. (2020), where a general dictionary with just the names of prominent political parties was used.
To construct definitions for terms like “Lockdown,” we turned to the keywords extracted from Nexis Uni. Additionally, for nuanced terminology related to political parties, such as alternative descriptors for the BJP, we leveraged terms derived from extensive news reporting spanning various periods, thus enriching the depth and contextual accuracy of our dictionary.
The dictionary was manually formulated and two coders were involved in making the dictionary, along with the authors. We have selected news articles randomly spanning over the entire time period for subjective scrutiny. On the basis of our scrutiny, we have included the following proper and common nouns in the dictionary created below.
