Radical right populism and supporter attitudes

PRR parties like the French Rassemblement National (RN – National Rally) and the Lega (League) in Italy are said to share a fundamental set of ideas, namely nativism, authoritarianism and populism (Mudde Reference Mudde2019). As Cas Mudde (Reference Mudde2019: 22) explains, nativism ‘holds that states should be inhabited exclusively by members of the native group (the nation) and that non-native (or alien) elements are fundamentally threatening to the homogeneous nation-state’. In Europe, non-natives for the PRR tend to be immigrants, especially in recent decades those from Muslim countries; but they can also include minority groups of different ethnicities that have been long present within a country, such as Romany communities in the case of Hungary (Pirro Reference Pirro2015). The nativism of the PRR is thus a combination of anti-immigrant attitudes and ethnocentrism. Regarding authoritarianism, we follow Bart Bonikowski (Reference Bonikowski2017: 189–190), who characterizes it as a ‘style of governance that attempts to circumvent the rule of law and democratic norms in favour of centralized authority and limited political freedom’. Nativism and authoritarianism are complemented in the PRR by a populist worldview that considers the beliefs, prosperity and sovereignty of a ‘virtuous’ and ‘homogenous’ people to be under attack from ‘bad’ elites, with the latter usually including a mixture of other political forces, the media, big business, the judiciary, bureaucrats, international organizations, academics and so on (Albertazzi and McDonnell Reference Albertazzi and McDonnell2015: 4–6). Invoking a sense of imminent crisis, radical right populists present themselves as the sole defenders of a sole true people and, indeed, of national democracy itself. However, while populism is not anti-democratic per se, it is at odds with the principles of liberal democracy (Albertazzi and McDonnell Reference Albertazzi and McDonnell2015; Mudde Reference Mudde2019).

Since Narendra Modi became party leader, the BJP has increasingly been seen not only as a Hindu nationalist party, but as a PRR one (McDonnell and Cabrera Reference McDonnell and Cabrera2019; Varshney Reference Varshney and Jayal2019). The BJP was founded in 1980 as the political arm of the Sangh Parivar, an umbrella of organizations committed to Hindu nationalism, or Hindutva, which are active in a wide range of sectors, from education to social welfare (Andersen and Damle Reference Andersen and Damle2019; Sarkar Reference Sarkar2021). The most prominent of these organizations, the Rashtriya Swayamsevak Sangh (RSS – National Volunteer Organization), can be considered not only the ideological parent of the BJP, but an essential source of the party's personnel, since almost the entire leadership and cadre base of the BJP come from the RSS (Jha Reference Jha2017: 131). Until the 1980s, the Hindu nationalist project was rooted in an elitist philosophy, and this was reflected in the fact that both leaders and members of the Sangh Parivar were predominantly from upper-caste backgrounds (Thachil Reference Thachil2014: 41). With the increasing mobilization of lower castes by regional caste-based parties like the Bahujan Samaj Party (BSP), however, the BJP realized the need to expand its electoral base beyond its traditional core of supporters, and to mobilize the Hindu population on the basis of a pan-Hindu identity (Thachil Reference Thachil2014). This strategy of electoral expansion began in the 1980s, but under Modi reached unprecedented success, with the BJP making substantial gains among Other Backward Classes (OBCs), Scheduled Castes (SCs) and Scheduled Tribes (STs) at the 2014 and 2019 national elections (Chhibber and Verma Reference Chhibber and Verma2019).Footnote ² The triumph of the BJP under Modi's leadership has been marked by the party's radicalization further to the right, and its embrace of the three core pillars of PRR ideology: populism, nativism and authoritarianism (Chatterji et al. Reference Chatterji, Hansen, Jaffrelot, Chatterji, Hansen and Jaffrelot2019; McDonnell and Cabrera Reference McDonnell and Cabrera2019).Footnote ³

Initially, scholars associated populism specifically with Modi rather than with the party, by focusing on his campaigning style (Chakravartty and Roy Reference Chakravartty and Roy2016; Jaffrelot Reference Jaffrelot2015) and mobilization strategies (Kenny Reference Kenny2017). However, there is now widespread recognition that the BJP under Modi promotes a set of ideas that must be understood within the context of the global right-wing populist wave (Chacko and Jayasuriya Reference Chacko and Jayasuriya2018; McDonnell and Cabrera Reference McDonnell and Cabrera2019; Sircar Reference Sircar2020). The party conceives of ‘the people’ in cultural terms as the Hindu community (the nation's religious majority) and it highlights the differences between Hindus and Muslims while underplaying caste divisions within Hindu society (Varshney Reference Varshney and Jayal2019). This framing has helped the BJP to de-emphasize its previous image as an upper-caste party mainly confined to northern India and appeal to a much wider social base (Chhibber and Verma Reference Chhibber and Verma2019; Jha Reference Jha2017; Kumar Reference Kumar2020).

In the BJP's worldview, Hindus are portrayed as being oppressed by the presence of ‘dangerous others’ within Indian society, especially Muslims. It claims that the latter not only espouse values that are incompatible with those of the true Indian people, but also pose a series of threats to the Indian nation, ranging from demographic ones concerning their increasing population to sexual ones, with Muslim men depicted as predatory and violent figures (Chatterji Reference Chatterji, Chatterji, Hansen and Jaffrelot2019). In addition to being menaced from below, the people are cast as under siege from above by political, cultural, media and financial elites, culpable of being self-interested and detached from Hindu identity and traditions (McDonnell and Cabrera Reference McDonnell and Cabrera2019). Moreover, the elites from the BJP's main opponent, the Indian National Congress (traditionally India's dominant party), are accused of favouring Muslims over Hindus by providing special benefits in exchange for electoral support. In order to counter these threats to the people and the expression of their will, the BJP has adopted authoritarian measures since taking power in 2014 that have weakened the pillars of India's liberal democracy, for instance by labelling protests as ‘anti-national’ and violently suppressing them, imposing new restrictions on NGOs, and interfering with the independence of the judiciary and the media (Gudavarthy Reference Gudavarthy2018; Varshney Reference Varshney and Jayal2019).

In this study we investigate whether the distinct PRR supply of the BJP is matched by a similarly distinct PRR demand among its electorate. Demand and supply of course interact in determining support for PRR parties. As Frank Mols and Jolanda Jetten (Reference Mols and Jetten2020) show, the latter supply a series of frames and narratives that derive from, and further fuel, a range of grievances already present within society. This loop mechanism is often enhanced by a leader who can create ‘a sense of shared victimhood’ among the people (Mols and Jetten Reference Mols and Jetten2020: 9). In the case of the BJP, as Nitasha Kaul (Reference Kaul2017) writes, Modi presented himself as a national leader who could take the Hindu people back to a glorious past. Furthermore, his popularity is also a product of the proselytizing work of the RSS, and the unprecedented amount of campaign spending that the BJP invested in vote mobilizers and media to promote his image (Chhibber and Ostermann Reference Chhibber and Ostermann2014; Jaffrelot Reference Jaffrelot2015; Kaul Reference Kaul2017). While we recognize the importance that such a solid PRR supply can have in determining support for the BJP, in this article we focus on a more overlooked, demand-side, aspect of the BJP's success, namely the prevalence of PRR attitudes in India and their role in determining support for the party. In other words, we ask: Are BJP supporters characterized by populist, nativist and authoritarian attitudes that distinguish them from supporters of other Indian parties?

Research in Europe and elsewhere has shown that support for PRR parties and leaders is underpinned by populist and nativist attitudes among the citizenry, but less so by authoritarian attitudes. While the recent study by Bruno Castanho Silva et al. (Reference Castanho Silva, Fuks and Ryô Tamaki2022) found that populist attitudes do not drive support for Jair Bolsonaro in Brazil, this is very much an exception. In fact, populist attitudes characterize supporters of populist parties and leaders in an array of countries, such as the Netherlands, the US, Greece and Chile (Akkerman et al. Reference Akkerman, Mudde and Zaslove2014; Hawkins et al. Reference Hawkins, Rovira Kaltwasser and Andreadis2020; Oliver and Rahn Reference Oliver and Rahn2016). As regards nativism, scholars have focused mostly on anti-immigration feelings among PRR supporters rather than a wider range of nativist positions, thus overlooking the dimension of ethnocentrism. This likely reflects the fact that the main ‘other’ for most European PRR parties are immigrants (particularly from Muslim countries) rather than long-standing ethnic minorities within countries. That aside, comparative studies across several European countries have shown that anti-immigration attitudes are systematically associated with support for a PRR party (Arzheimer and Carter Reference Arzheimer and Carter2009; Lubbers et al. Reference Lubbers, Gijsberts and Scheepers2002; Werts et al. Reference Werts, Scheepers and Lubbers2012). The evidence for authoritarianism, however, is not as straightforward, with studies coming to different conclusions about the role of these attitudes in explaining PRR support (Donovan Reference Donovan2019; Dunn Reference Dunn2015; Lubbers and Coenders Reference Lubbers and Coenders2016; Werts et al. Reference Werts, Scheepers and Lubbers2012). While most analyses have addressed only one or two sets of PRR attitudes, we follow the approach of Agnes Akkerman et al. (Reference Akkerman, Zaslove and Spruyt2017) and Steven van Hauwaert and Stijn van Kessel (Reference Van Hauwaert and Van Kessel2018) by looking at the combination of all three ideological pillars. This is also because, as Matthijs Rooduijn (Reference Rooduijn2014) shows in the case of the Netherlands, the three key ideological features of PRR parties – populism, nativism and authoritarianism – can form a coherent set of attitudes among the public. He concludes therefore that populist, nativist and authoritarian attitudes ‘should be studied in conjunction with each other in order to explain PRR voting behaviour’ rather than separately, as had been done until then (Rooduijn Reference Rooduijn2014: 81).

As noted in the introduction, there is reason to think there may be differences between the mostly European cases studied to date and the BJP. Researchers have shown that support for PRR parties is as motivated by ideological concerns as support for any other party (Van Der Brug et al. Reference Van der Brug, Fennema and Tillie2000). However, the characterization of India as a ‘patronage democracy’ (Chandra Reference Chandra2004; Kenny Reference Kenny2017) has translated into the prevailing wisdom that contemporary Indian politics is primarily non-ideological and is instead dominated by pragmatism, clientelism and personalism (Chhibber and Verma Reference Chhibber and Verma2018). Scholars have warned that we should ‘not overemphasize the role of ideology in India's party politics’ (Suri Reference Suri, Suri and Vanaik2013: 240) because even if parties ‘do profess to stick to their party ideology … in their actual support they seem to be more pragmatic’ (Hasan Reference Hasan, Jayal and Mehta2010: 248). For these reasons, the political behaviour of Indian citizens is rarely described as being guided by ideological and programmatic motivations; rather, it is seen as the outcome of religious and caste considerations (Chandra Reference Chandra2004; Heath and Yadav Reference Heath, Yadav, Heath and Jeffery2010; Shah Reference Shah2004), aimed at accessing state and private resources through clientelist networks (Chatterjee Reference Chatterjee2004; Piliavsky Reference Piliavsky2014; Ziegfeld Reference Ziegfeld2016), and rooted in trust and feelings of reverence for a strong leader (Sircar Reference Sircar2020). Despite recent studies challenging this (for a review, see Auerbach et al. Reference Auerbach, Bussell and Chauchard2021), the dominant view remains that ‘far less “ideology” is observed in India than in consolidated Western democracies’ (Sircar Reference Sircar2020: 15).

The BJP thus provides an excellent case to investigate the global reach of PRR attitudes, and to assess whether European-derived theories about the PRR are equally applicable in a very different political and cultural context. Since we know that the three key ideological features of PRR supply are populism, nativism and authoritarianism, these give rise to our three hypotheses about PRR demand. The first concerns populist attitudes. In recent years, researchers have used surveys to investigate the extent to which people hold populist beliefs about political and societal issues. Most notably, Akkerman et al. (Reference Akkerman, Mudde and Zaslove2014) devised a scale of items to measure populist attitudes among the public. These items tap into the core dimensions of populism: people's sovereignty, antagonism towards the elites, and the Manichean division between ‘good’ and ‘evil’. Researchers have subsequently refined and tested Akkerman et al.'s scale to improve the measurement of individual populist support (Castanho Silva et al. Reference Castanho Silva, Jungkunz and Helbling2020; Van Hauwaert et al. Reference Van Hauwaert, Schimpf and Azevedo2020; Wuttke et al. Reference Wuttke, Schimpf and Schoen2020). Scholars have also focused on assessing whether populist attitudes among citizens translate into support for populist parties. Their findings, which are largely consistent across both single-case (Akkerman et al. Reference Akkerman, Mudde and Zaslove2014, Reference Akkerman, Zaslove and Spruyt2017; Oliver and Rahn 2016; Rooduijn Reference Rooduijn2014) and comparative studies (Hawkins et al. Reference Hawkins, Rovira Kaltwasser and Andreadis2020; Van Hauwaert and Van Kessel Reference Van Hauwaert and Van Kessel2018), show that citizens with stronger populist attitudes are more likely to support populist parties. Based on this research, we expect to find that:

The second component of the PRR worldview is nativism. This is the central ideological tenet of the PRR party family (Mudde Reference Mudde2019), and successful PRR parties are those that are able to mobilize grievances over non-native elements within society (Ivarsflaten Reference Ivarsflaten2008). Depending on whether ethnic minorities or immigrants constitute the main ‘non-native’ for a PRR party, nativism can be more driven by ethnocentrism or anti-immigrant attitudes. In Europe, where immigration is a constantly salient issue, negative attitudes towards immigrants are the strongest predictor of PRR support (Lubbers et al. Reference Lubbers, Gijsberts and Scheepers2002; Werts et al. Reference Werts, Scheepers and Lubbers2012). While research to date in Europe has focused on anti-immigration as the key indicator of nativist attitudes, our analysis will comprise both ethnocentrism and anti-immigration. This is especially important in India, given that, more than immigrants and refugees from Muslim-majority neighbouring countries, the principal ‘non-native’ for the BJP is the Muslim minority that has been present for centuries (McDonnell and Cabrera Reference McDonnell and Cabrera2019). Considering the relevance of nativism to the BJP's ideological platform, and its importance in explaining PRR support elsewhere, our second hypothesis therefore proposes that:

The third component of the PRR worldview is authoritarianism. Research to date has shown that authoritarian attitudes are not a consistent predictor of support for the PRR in the same way as nativism and populism are. For example, while Rooduijn (Reference Rooduijn2014) demonstrates how authoritarian attitudes in the Netherlands help explain support for the Partij voor de Vrijheid (PVV – Party for Freedom), Akkerman et al. (Reference Akkerman, Zaslove and Spruyt2017) arrive at the opposite conclusion: that authoritarianism does not distinguish PVV supporters. Likewise, Han Werts et al. (Reference Werts, Scheepers and Lubbers2012) find that authoritarianism is not a significant driver of PRR support in their analysis of 18 European countries, but Marcel Lubbers and Marcel Coenders (Reference Lubbers and Coenders2016) in their study of almost the same set of countries conclude that authoritarianism is in fact significant. This uncertainty regarding the role of authoritarianism is reinforced by Kris Dunn (Reference Dunn2015: 376), who shows that authoritarian attitudes underpin PRR support in only two of the five Western European states he examines (Denmark and Switzerland). In light of these mixed findings, it is hard to make predictions about authoritarian attitudes among BJP supporters with the same confidence as we did for nativism and populism. Moreover, the picture is further complicated by the fact that, according to a 2017 Pew Research Center report, many Indians hold authoritarian views, and support for an autocratic regime is higher there than in any other country that was surveyed (Stokes et al. Reference Stokes, Manevich and Chwe2017: 18).Footnote ⁴ Nonetheless, given that the theoretical literature has long affirmed the presence of authoritarianism in PRR party ideology, we (tentatively) opt for a hypothesis confirming this presence also on the demand side. We thus envisage that, even if authoritarian views are widespread in India, supporters of the BJP should hold even stronger ones.

Data and method

Our empirical analysis is based on an original survey conducted online with a sample of Indian respondents. Participants were recruited using a local contractor, which offered material incentives in exchange for participation.Footnote ⁵ Between August and September 2018, 1,526 respondents took the survey on an online platform that allowed them to access the questionnaire (provided in Hindi, or, on request, in English) through personal computers, tablets or smartphones. Given that less than half of Indians had internet access in 2018,Footnote ⁶ our methodology necessarily excludes an important section of the Indian population. Nevertheless, we are confident in our ability to extrapolate meaningfully from our sample. First, while the use of non-random samples such as samples from online panels is generally discouraged for descriptive inferences about a population, online samples are considered as adequate to study relationships between variables, and to analyse political ideology in particular (Baker et al. Reference Baker and Brick2013; Clifford et al. Reference Clifford, Jewell and Waggoner2015). Second, to ensure the representativeness of the sample, quotas for age, gender, region (rural/urban), education and religion were applied. The resulting sample contained considerable diversity in terms of sociodemographic factors, as shown in Table A1 of the Online Appendix, where we compare the applied population quotas and sample proportions for these five variables. For example, our sample includes substantial numbers of respondents who are typically underrepresented in online samples, such as women (45.3%), individuals with low education (49.5% have attained elementary education or less), and older respondents (15.3% of the sample are at least 55 years old). These figures show that our sample presents sufficient heterogeneity to explore our hypotheses.

To gauge support for the BJP, our dependent variable, we use an indicator of partisanship that captures closeness to the party, for which we ask respondents to say if they feel not close at all, not very close, close, or very close to the BJP. We opted for partisanship rather than vote for the BJP given the increasing tendency among Indian electors to engage in split-ticket voting by favouring the BJP at the national level, but not necessarily at the state one (Aiyar and Sircar Reference Aiyar and Sircar2020). Overall, support for the BJP is quite high in our sample, as 45% of respondents feel very close to the party, 29.3% feel close, 16.3% not very close and the remaining 9.5% not close at all.

To measure populism, we follow Van Hauwaert and Van Kessel (Reference Van Hauwaert and Van Kessel2018) and develop an index of populist attitudes based on seven items designed to measure support for a populist worldview. While some, such as Alexander Wuttke et al. (Reference Wuttke, Schimpf and Schoen2020), have proposed to conceptualize populism as a non-compensatory, multidimensional construct, we adopted a well-tested approach in which populism can be conceptualized as anti-elitism, combined with a strong preference for popular sovereignty. The statements in our survey items enunciate populist tropes such as the contraposition between a virtuous people and corrupt elites, the idea that decisions should be taken by the people rather than professional politicians, and a distaste for political compromise. Respondents express their agreement or disagreement with each item using a Likert scale (1 = strongly disagree; 2 = disagree; 3 = neither agree nor disagree; 4 = agree; 5 = strongly agree). An index is then built by calculating the mean value of these responses.Footnote ⁷ This scale, which has already been applied in studies of populism in Asia (Fossati and Mietzner Reference Fossati and Mietzner2019), is a reliable indicator of populist attitudes in our sample (α = 0.78). Most respondents could be described as populist in their worldview, as the median value for the index is 4.43 and 77.9% score an index value between 4 and 5.

As discussed in the previous section, we understand nativism as consisting of a combination of ethnocentric attitudes and anti-immigrant sentiments. For ethnocentrism, we ask about agreement with the following statements: ‘Other cultures should be more like my culture’ and ‘I dislike interacting with people from different cultures’ (Bizumic et al. Reference Bizumic, Monaghan and Priest2020). For anti-immigrant attitudes, the statements are: ‘Immigrants increase crime rates’ and ‘Immigrants take jobs away from people who were born in India’ (Lubbers and Coenders Reference Lubbers and Coenders2016). We compute the arithmetic mean of these four items to build an index similar to our indicator of populism, where higher index scores denote a higher level of nativism. This measurement also presents high levels of reliability (α = 0.79), and respondents display overall lower levels of nativism than populism with a median index value of 3.75. Given the BJP's Hindutva ideological platform, we also thought it useful to measure support for Hindu nationalism among respondents. To do so, we use an indicator regularly employed in research on national identity (Reeskens and Hooghe Reference Reeskens and Hooghe2010), which asks how important it is to be a Hindu to be ‘truly Indian’. Of our sample, 70% describes this factor as somewhat or very important to be Indian.Footnote ⁸ Finally, we operationalize authoritarianism with a commonly used question in survey research (Donovan Reference Donovan2019), namely one that asks for agreement or disagreement that ‘India needs a strong leader who can disregard the law for the good of the people.’ In accordance with previous research, which has suggested that authoritarian values are widespread among the Indian public (Stokes et al. Reference Stokes, Manevich and Chwe2017), our respondents largely support this idea, as 55.3% of the sample strongly agree with this statement.

We also include in our models a series of sociodemographic variables that may be associated with political attitudes and behaviour, namely age, gender, religion (Hindu versus minority), place of residence (rural versus urban), education and income (measured with ordinal variables with four education levels and five income brackets). Table 1 reports descriptive statistics for all the measures we use, and a correlation matrix is included in the Online Appendix (Table A3).

Table 1. Descriptive Statistics

In the next section, we estimate a series of ordered logistic models in which closeness to the BJP is a function of the indexes of populist, nativist and authoritarian attitudes. In each model, we investigate how robust the findings are to inclusion or exclusion of the various covariates. Missing values are handled with listwise deletion.

BJP supporter attitudes

Table 2 reports estimation results for a set of models in which our indicator of ‘closeness’ measures support for the BJP.Footnote ⁹ The four models differ in the covariates they include. All of them are estimated with the sociodemographic variables: the first includes the index of populist attitudes, the second the index of nativism and our measure for Hindu nationalism, the third our measure of authoritarianism, and the fourth estimates coefficients for all the ideological pillars combined. Table 2 shows that populist attitudes are estimated to have a positive and significant relationship with identification with the BJP in all models in which the index is included, although the magnitude of the coefficient varies with model specification. Using the full model (Model 4) as a basis for calculation, a respondent with strong populist views has a higher expected probability of feeling very close to the BJP than one scoring lower levels of populist attitudes. For example, a populism score of 4.5 is associated with an expected probability of 45.6%, while a score of 3 has an expected probability of 28.9%.

Table 2. Drivers of Closeness to the BJP

Notes: Standard errors in parentheses. *** p < 0.001, ** p < 0.01, * p < 0.05.

A similar pattern can be identified with respect to nativism, also a strong driver of closeness to the BJP in all models. Using again the illustration of how different index values are associated with identification with this party, the expected probability of feeling very close to the BJP rises from 34.6% in a respondent with a nativism index score of 3 to 53.5% for an index value of 4.5. Nativism thus appears to be a slightly more powerful driver of BJP support than populism. Our results are similar for Hindu nationalism. For instance, a respondent who believes Hinduism is not at all important for being Indian has an estimated probability of feeling very close to the BJP of 25.1%, but this figure rises to 53.1% among those who say being Hindu is very important to being Indian. Individuals who feel close to the BJP can therefore be described as significantly more nativist, more populist and more Hindu nationalist than those who do not feel close to the party. The fact that Hindu respondents are also substantially more likely to report feeling close to this party (more precisely, their estimated probability of feeling very close to the BJP is 11.6% higher than for non-Hindus) is related to the high levels of nativism in the BJP support base, and resonates with the polarization patterns between Hindus and religious minorities discussed above.

A different picture, however, emerges for authoritarian values, which do not appear to be systematically and consistently linked with closeness to the BJP. In Model 3, which only includes demographic covariates and the authoritarianism measure, the latter is significant and positively signed, suggesting that higher levels of authoritarian attitudes are indeed more likely among those who feel close or very close to the BJP. Nevertheless, the estimated coefficient in Model 4, where all PRR attitudes are included, is much more modest in magnitude and not significant at conventional levels. Based on this evidence, we can conclude that BJP supporters are no more likely to present strong authoritarian attitudes than those who do not report feeling close to the party. In short, they are distinctly nativist and populist, endorsing two of the three ideological orientations that are typically associated with PRR parties. At the same time, however, authoritarianism does not seem to be an exclusive feature of BJP supporters, but is instead more widely distributed among Indian citizens.

The coefficient plot reported in Figure 1, based on the full model, summarizes the estimation results and displays the substantial difference between the effect of nativist and populist attitudes and that of authoritarianism. The plotted coefficients also show the estimated association between BJP support and the various demographic indicators we have included in the model. In addition to religion, the only other covariate that displays a significant association with the dependent variable is education, as its estimated coefficients indicate that respondents with higher educational levels are systematically less likely to feel close to the BJP. The probability of feeling very close to the BJP is estimated at 56.1% among respondents who have attained primary education or less, 32.2% for those with middle-school education, 30.2% among respondents who have finished high school and 26.5% for college-educated respondents. BJP supporters are therefore more likely to be individuals with lower educational attainment, which reflects the party's expansion of its social base under Modi's leadership (Chhibber and Verma Reference Chhibber and Verma2019). The negative relationship between education levels and BJP support does not, however, contradict the fact that the BJP is popular among upper-middle and upper classes (Fernandes Reference Fernandes and Jacobsen2015). On the contrary, and together with the finding that income does not appear to have a linear relationship with support for the party, our results reinforce the idea that the Modi-led BJP's narrative appeals both to those who directly benefit from economic liberalization, and to those who struggle with it (Fernandes Reference Fernandes and Jacobsen2015: 239). This is at least partially thanks to Modi's ability to speak to different constituencies (Jha Reference Jha2017; Kaul Reference Kaul2017).

Figure 1. Coefficient Plot of Factors Associated with BJP Support

As for the other covariates, age and gender are not associated with BJP closeness and are not significant drivers of support for the BJP. These findings are not overly surprising given that an increasing number of women have voted for the BJP since Modi took over, as have young people. Or, as Pradeep Chhibber and Rahul Verma (Reference Chhibber and Verma2019: 139) put it, ‘the traditional BJP voter was more likely to be male and older, whereas the newer BJP voter is equally likely to be female and young’.

To probe the robustness of our findings, in the Online Appendix we have re-estimated the models reported in Table 2 with a measure of support for Narendra Modi, who has led the BJP's full embrace of radical right populism and guided the party to unprecedented levels of success. Specifically, we have used a five-point Likert scale to measure if respondents strongly disagree, somewhat disagree, neither agree nor disagree, somewhat agree, or strongly agree that Modi is the right person to lead India. When the models are estimated with this dependent variable, all of the substantive results reported above hold (see Online Appendix, Table A4). This test reassures us that our findings are not sensitive to specific operationalizations of the dependent variable.