Questions over redistribution and social justice politics versus religious fault-lines seem front and center once again in India’s ongoing general elections. How is the ruling Bharatiya Janata Party—which has historically been ambivalent about social justice politics—mobilizing poorer voters, especially among the upper castes? What role does “social status” play in how voters choose their candidates? And when do anti-redistributive politics become salient?
In the fourth interview of the CASI Election Conversations 2024, CASI Consulting Editor Rohan Venkat speaks to Pavithra Suryanarayan (Assistant Professor in the Government Department at the London School of Economics and Political Science) about her research into the awakening of “social status” as an identity category, when social justice politics leads to a backlash that ends up hollowing out the state, and what other big questions about Indian politics she would like to see studied.
Rohan: Could you tell us what you’ve been seeking to understand in your research?
Pavithra: I broadly work on three topics. First, I study how people's identities shape vote choice and their attitudes toward redistribution. Second, I study the politics of building state capacity. The third is an area of research that I have developed with a co-author, Francesca Jensenius, which focuses on party organization and party systems in the Indian states. I believe our conversation today will focus a lot on the first research agenda, which is how do people's identities shape their worldview and by extension, their political behavior?
Rohan: Quoting from one of your other papers, “scholars of the world’s most populous democracy invariably agree that elements of identity are central to Indian politics, but there have been few efforts to assess empirically which dimensions of identity are most important to understanding electoral choice, or to understand how the economic attributes of groups might be related to the salience of group identity in elections.” I want to get a sense of what is already out there in the field on these questions before we get to your work.
Pavithra: India is a country teeming with multiple identities—religion, language, caste. Of these identities, caste is what I focus on. On an everyday level, people don't just experience their caste as upper caste or lower caste, they also experience it as a localized jati or biradari, which then shapes a whole range of socio-economic outcomes—what jobs they do, who they marry, what they eat, how they dress, to even the dialect they speak. Caste is interesting because it has many attributes. In American political science, there was a recent paper called “Race as a Bundle of Sticks,” because race in the American context is a bundle of many attributes. We can think about caste in a similar way. Caste endows a person social rank, occupation, phenotype, a linguistic dialect, and caste can be a bundle of cultural or religious rituals you practice as a consequence of the local jati that you inhabit.
If we think about caste as an identity that matters to politics, we have to ask, “which attribute of caste is playing on people's views on what they want from politics?" I've tried to tease out a few of those threads, most notably how the rank that people inherit from caste shapes political behavior.
Rohan: We’ve always understood that caste is important, with the old cliché about not casting your vote, but “voting your caste.” What you’re trying to do is get to what, within caste, is important. What do you mean by “rank” here?
Pavithra: Now we come to the part of my research that tries to put some structure and shape to how to think about this. I'm going to use the phrase “social status,” which is quite loaded because it can mean very different things to different people depending on what trajectory of social science research they follow. But here, when I say social status, I mean an inherited rank that comes through descent. Social status is relatively sticky; you cannot get rid of that inherited status that you have received as you go through your life. You may become richer or poorer but your inherited rank stays with you. What makes this a relatively new conceptualization of social status is that rank is not an attribute, it is the constitutive attribute of the identity. If you lose the social status distinction, the identity category ceases to exist.
In my work, I theorize caste as a social status identity, which is different from viewing it as an ethnic category or a class category. We have extensive research on caste as an ethnic category and how it shapes politics, voting, and political parties. Much of this research, however, doesn’t consider what constitutive attribute of a particular caste might be the key factor in increasing caste salience.
Thinking about caste as a social status identity, where rank is the chief constitutive attribute of the identity, allows us to then think about when, why, and for which castes social status emerges as a salient identity cleavage in politics. It allows us to more clearly understand the factors that come into people's utility functions when they think about caste and politics.
Rohan: To clarify for those unfamiliar with social science terminology, “rank” here is the perceived idea of where you are situated in society on a hierarchical scale.
Pavithra: That's right. Thank you for pointing that out, because it might not just be for people who don't understand social science. There might be people who don't understand how caste as an identity is structured. Historically, a central organizing feature of the caste system was hierarchy. We can think about it as the classic Varna categories of Brahmin, Kshatriya, Shudra, and so on. Or we can just think about it in contemporary terms as the agglomerations that social scientists in India use as upper caste, peasant proprietary caste, the upper creamy layer other backward classes (OBCs), lower OBCs, scheduled caste, scheduled tribes (SC/ST), and so on.
When you face challenges to your social status—ie, your inherited rank in the caste system—what you're fighting tooth and nail for is to maintain rank as the organizing principle in a society. What you're, in fact, afraid of is a state of the world where caste is merely ethnicity or caste is merely class, which takes away the utility that people get from inheriting rank in a caste system.
Rohan: I wanted to focus on the paper titled, “When do the poor vote for the right-wing and why: Hierarchy and vote choice in the Indian states.” Can you tell us what you set out to do in the paper and what it finds?
Pavithra: In this paper, I do something that we don't regularly see in Indian politics. I theorize that upper caste Brahmins hold a distinct set of motivations for politics. There is a graph in that paper that caught my attention when I first made it using survey data from 2004 National Election Surveys. We don't have much information on castes enumeration or the economic location of castes in contemporary India because we haven't done a caste census since independence. The NES surveys that document jati-level data are some of the best measures of the economic and social location of caste groups across the Indian state.
If we look at the distribution of wealth across India's caste groups, we see something intriguing. One, that Brahmins, who are India's upper castes, are more heterogeneous in their economic profiles than scheduled castes and tribes. While on average, they tend to be the wealthiest group across the country, there are a number of poor Brahmins in India. What also caught my attention is that intra-Brahmin inequality looks a lot like intra-upper OBC or intra-other upper caste inequality. I wanted to understand: do poor Brahmins then think about politics in the same way as poor backward castes?
Here, what do I mean by politics? I mean redistributive politics—what do they want from their government in terms of redistribution? Are poor Brahmins like other poor backward castes in that they want the government to do more welfare, more tax-and-transfer, and to do more in poverty alleviation? Or are poor Brahmins similar to wealthier Brahmins in that they hold more circumspect redistributive preferences where they want the government to limit its role in the economy and get out of managing welfare? What grew out of this chart was an interest in understanding India's upper castes, particularly Brahmins.
This is quite new because historically, much of social science work has gone into understanding how lower castes, particularly marginal castes like SC/ST and lower OBCs, have made claims on the state or on politics. I'm focusing the attention on upper castes and making the claim that this group has expressed very distinctive political preferences in different time periods in history.
Rohan: What do you find? What are poor Brahmins expressing in terms of preference?
Pavithra: This particular paper makes use of the time period before and after Prime Minister V.P. Singh's speech around implementing the recommendations of the Mandal Commission (introducing reservations for OBCs in government jobs and higher education). It argues that there was a shift in voting away from the Congress and toward the BJP in the state elections that were held right after the Mandal announcement in 1990. The paper argues that the Mandal speech by Prime Minister Singh awakened a social status identity that was somewhat dormant in upper castes in the previous period, because government policy in India until that point hadn't overtly gone into arenas that disturbed upper caste privilege. And here, a central feature of upper caste, particularly Brahmin privilege, historically, has been their control of education. It's a central aspect of Brahminical identity, intertwined with their role as priests. The prospect of desegregating higher education institutions, and by extension, the bureaucracy that comes with having held access to education historically, was interpreted as a status threat amongst India's upper caste Brahmins.
This catalyzed a shift away from supporting parties like the Congress in a previous era and toward the BJP, which, in this time period, started to explicitly speak the language of de-bureaucratizing and anti-affirmative action policies. And so, this moment allows you to understand how an identity like social status, which derives from rank, shapes people's voting preferences.
If this argument is right, would we see the effects most sharply in some places over others? This involved actually going back to the most comprehensive census we have of local caste groupings and caste inequality around status, which is the 1931 census. I digitized the 1931 census data at the taluk-level in order to then be able to make claims that that same underlying structure of caste and status inequality suddenly came alive as a key predictor of voting after the Mandal announcement precisely in places where that inequality around rank was greater.
Rohan: What you discover is that, based on this data from the 1920s and 1930s, Brahmin-dominated spaces corresponded to preferences for the BJP after reservations were announced in 1990. And this holds up specifically for poor Brahmins?
Pavithra: Part one of the paper allowed me to first illustrate that places which had more Brahmin dominance [in 1930] had a greater rise in voting for the BJP in the period after the Mandal announcement. The second part of the paper gets into the individual level dynamics of who these people are who support the BJP. I used data from the 2004 National Election Surveys. I was able to show that upper castes, particularly Brahmins, who lived in historically Brahmin-dominated areas, were more likely to vote for the BJP. Importantly, poor Brahmins living in Brahmin-dominated areas were more likely to vote for the BJP.
I also explored a series of questions on redistributive politics, which ask voters things like, is it the government's role to care for poverty? I find that poor Brahmins are more anti-redistributive if they live in historically Brahmin-dominated areas. There's something about the context where rank and social status is more salient that it turns the poor members of a high rank group against state-led welfare and redistribution. It links the support for the BJP to the evolution of anti-redistributive politics amongst high status groups.
Rohan: Since this is also the time of the Ayodhya movement and the Babri Masjid demolition, couldn’t that explain the upper-caste support for the BJP? How do you conclude that it was specifically a response to Mandal?
Pavithra: In the paper, just like I created measures of Brahmin dominance, I also create measures of Muslim size and population. I rely on the intuitions of other scholars of ethnic politics who have argued that, "you might perceive the threat of out-groups more if you're more equally sized to them." If it is Babri Masjid and the rise of Hindutva politics that was doing this and not Mandal, then either controlling for size of Muslims in a constituency should take away the effects of the historic Brahmin dominance measure or the measures should work through Brahmin dominance in that, what they're doing is picking up more upper caste support for Hindutva politics. And neither of those two things are what happens when I examine the data.
The data shows there is an effect of Muslim populations in the assembly constituencies in rallying support for the BJP, but that doesn't take away the independent role of Brahmin dominance in those districts. Second, the effect of the Hindutva ideology seems to be greater in low-Brahmin-dominant areas than high-Brahmin-dominant areas, which means two things are happening simultaneously in India in this moment. One, the reverberating effects of Mandal that are shaping the behavior of India's upper castes and the effects of mobilization around a Hindu identity, which might be allowing the BJP to seek votes from non-upper caste groups against their political competitors. The strongest version of my claim would be that Hindutva itself was, in some ways, a reaction to the fallout of Mandal, that the BJP needed a strategy that could both bring Brahmins and upper castes on board, while also rallying non-upper castes into the fold. But a different empirical paper needs to be written to push that claim.
Rohan: The BJP no longer seems quite as anti-reservations or anti-redistributive in the way it was in the 1990s…
Pavithra: I will say two things. I still think the BJP is one of the few parties that does not believe in affirmative action. It doesn't overtly act on it, but if you engage with its party platforms and what its leaders have been saying, it has a very non-affirmative action policy bench. And that has stayed quite consistent over the last 40 years, since Mandal. The other thing I will say is that we have to ask, what sort of redistribution is the current BJP doing? A central argument in my paper and my ongoing book project is that broad strokes redistribution may not threaten upper caste status identity. This is why India's upper caste were quite comfortable with the redistributive policies of the Grand Old Party Congress.
The kinds of policies that upset upper caste enclaves are attempts to genuinely integrate institutions. This is why Mandal was such a pivotal moment, because here was a prime minister who said, "There are limitations to political power. Political power means nothing unless institutions themselves are run by lower castes." Explicit in Prime Minister V.P. Singh's speech was this idea that lower castes not only needed political representation, they needed to control the state. They genuinely needed to be part of key enclaves of upper caste dominance.
And you don't see the BJP doing much of that. The BJP in its current form is toeing the line between appeasing its upper caste interests of institutional control while also providing redistribution of the form that doesn't disturb these enclaves. The current BJP is a good manifestation of how an ideal upper caste party wants to function. And this is also reflected in findings by Tariq Thachil who discusses how BJP and its allied organizations provide social welfare to marginal groups while allowing policy and institutions to remain in the hands of their upper caste core constituents.
Rohan: There’s a complex interplay there, given that the BJP always embraced anti-reservation thought, but also has had to embrace other strains of political culture, so much that its big move ahead of the 2019 elections was to bring in reservations for upper castes in the form of the EWS quota. To come back to the paper and book project, what are the implications of your findings, not just for India but this research at large?
Pavithra: The book project develops the idea that if you want to understand redistributive politics in some countries, you need to take two things seriously. The first is that the contestation in politics doesn’t only hinge on material redistribution, it can hinge on rank or social status.
The second is that we focus overtly on representation in the political arena, which is fighting elections and putting an MLA or an MP into office, and then we say, “job done." But in fact, the arena that might truly matter is the state itself. Control of the bureaucracy and bureaucratic capacity might be important to certain groups in a society.
Historically in many places like India, the United States, South Africa, etc., the state has been key to not only providing security and taxes, but has been key to social organization. In these countries, the state was central in allowing segregation to continue. Integrating the state can create a backlash that reverberates in politics.
I have a paper on the American South on how the end of the Civil War produced this backlash as African Americans were emancipated and made claims on redistribution, and many of them made claims through becoming part of the state. African Americans became not just voters and representatives, they became census enumerators and local bureaucrats. What followed was a period of backlash against the state, which led to weakened state capacity and a deliberate attempt to keep the state weak in order to limit its welfare ability.
The book spells out how the increased salience of a social status identity leads to anti-redistributive politics. We observe anti-redistributive politics in three ways. First, at the micro level, it lowers individual level support for redistribution amongst high status groups. Second, at a meso level, it weakens state capacity, particularly bureaucratic capacity. Third, at a macro level, it changes coalitions of voters in the political landscape, making it possible now for right-wing parties to gain new constituents who are worried about protecting their social status and not just their income or their ethnicity.
Rohan: How do we see that backlash in the Indian context?
Pavithra: I use three different methods to get through this micro, meso, macro. At the micro level, we have just completed a series of survey experiments in Uttar Pradesh with a colleague, Simon Chauchard, where we examine upper caste versus middle caste—which is upper OBC—preferences for redistribution once they're experimentally informed about who controls the state. Tentatively, what we find is that upper castes turn more anti-redistributive when exposed to information about a more integrated state. We see no such trend amongst either the wealthy or the poor upper OBCs in the survey experiment.
For meso-level institutional findings, I go back a hundred years to what happens to tax collections and tax capacity in colonial India after the Montagu-Chelmsford Reforms. I focus on two provinces in colonial India—Madras and Bombay—where the status threat was very high because in the pre-Montagu world, the bureaucracies of these provinces were almost entirely dominated by upper caste Brahmins. I show that Brahmin-dominant areas experienced a greater weakening of taxation and tax collections after Montagu-Chelmsford Reforms are announced because again, the threat of integrating the state weighs very heavily on upper-castes, who, in these places, have historically consumed key goods like education that were central to their caste identity.
Historic upper caste dominance, particularly Brahmin dominance, is associated with an anti-tax backlash, even while controlling for other factors that could explain the same phenomenon like land-based fights and land inequality or simply ethnic fragmentation or ethnic competition at the local level. And then you have the macro level findings, which is where we started this conversation—the BJP paper that shows the same thing in contemporary India before and after Mandal.
Thinking about these at the micro, the meso, and the macro gives us a fuller picture of how the same impulses shape an anti-redistributive politics, but only when the status is threatened. These are groups that were very pro-bureaucracy, very pro-redistribution, much more comfortable in welfarist politics. But they shift to an anti-redistributive mindset when their status is threatened. And that's what the survey experiment shows, that's what that moment of Montagu-Chelmsford Reforms shows, and then again, Mandal. And I'm hard-pressed to think if this is totally wrong, what alternative explanation could explain all three of those findings in exactly the same direction or way.
Rohan: There is often this idea that poorer voters who support anti-redistributive parties are making irrational choices—voting against their interest by going against redistributive politics. This could work to explain that, and it also made me think immediately about whether it explains efforts to make white communities in America, or Hindus in general, feel threatened in their rank.
Pavithra: In my book manuscript and my papers, I push back against the notion that poor whites or upper castes are “irrational.” Instead, what they are attempting to maximize are a rank-based set of goods. Losing rank means they no longer belong to that exclusive club that allows them a whole range of material goods, not just psychological wages, as Du Bois called, "wages of whiteness." Yes, there are some psychological or psychic goods that people get, but they also get a very specific set of goods that has been so important to reifying rank in the first place. This is why I started with education because this was the domain of Brahmins for centuries, and one that allowed them to then become such a key part of the state through both the pre-colonial and colonial era.
This particular good—education—was key to the rank in social life. Your local priest or your teacher is the one with the words, the one who knows how to read the religious text and gets a great deal of social status bestowed on them as a function of that, and then becomes a key player within the state and gets to set policy and rules again, based on that rank attribute of education being so central to Brahminical social status. What I want to say is that it's rational for voters in different contexts to try and defend these segregated goods because it's a form of inheritance that they give to their children, this rank, and they fight tooth and nail to maintain it and prevent the world becoming about an ethnic group one versus ethnic group two or rich versus poor dynamic. This defense of rank, it's fully rational if what you're trying to maximize here is your social status.
Then comes the second question—“When we see Hindu-Muslim dynamics in contemporary India, would this be similar?" Maybe. I don't want to do too much concept-stretching. Maybe we are in a hundred-year process by which Muslims in India are being relegated to a certain social rank, which over time might start to resemble what we are seeing with upper versus lower castes in India. But I'm not comfortable making that conceptual stretch. I think a different analogy might work, and this is gender. Here too, you see status differences because of a function of birth, and is gender a form of rank, and can the concepts of defending status and rank apply to gender politics? Possibly. And this is something that's on my mind as I explore future research.
Rohan: Have there been critiques or responses to the work that you have found useful?
Pavithra: I think the first critique people make is that it's hard to think about status without its material underpinnings. How can you imagine that somebody has social status without also having wealth or some form of capital or income or sources of power that social status needs? I think this is why the study of the Indian caste system has been so illuminating. It allows me, both historically and in the contemporary moment, to find groups that are relatively evenly placed on their economic attributes, but differ on their birthplace rank and how it's here that you see the difference. That's a criticism I face a lot, and I think a lot about. The second criticism comes from ethnic politics scholars who say, "Well, you've just defined ethnic group A versus B, of which maybe what they're fighting over is rank. So, how is this different?"
And to them I say, well, rank is a very central and constitutive feature of the identity. This is not a case that just because Brahmins lose rank, they'll go on to just being happy about their cultural group and become one other ethnic group in India's many ethnic groups. No. I think here we need to understand what game it is that upper castes want to play versus what game it is that other groups are playing. And if they're not even playing the same game, you cannot expect that they will behave like the other group. And so, if one group wants to maximize wealth and income and the other group wants to maximize social status and rank, well, these are different games. That's another area for which I get criticism.
The third is the kind that comparative politics scholars give, which is, "Oh, is this all about caste? How far does this carry?" This is why it was important for me to think about, well, where else can I show this at play? I wrote the paper with Stephen White on the American South to show that actually the defense of rank carries very nicely to other contexts like America where the defining feature is race, not caste, but it's the same inherited social status that poor whites are defending. So, I think these are roughly my critics and I hope to keep doing work to rally to the challenge.
Rohan: Has it been difficult to make the case that the study of India cannot be translated or compared to the rest of the world?
Pavithra: Well, much less so now than when I first started my Ph.D. I remember I had to always prepare in every presentation a little slide about, "Why should you care after learning X about India?" And I think we are living in a moment where there are so many amazing social scientists across political science, economics, sociology, anthropology, demonstrating how the study of India's social, economic, and political structures is meaningful and valuable to compare to politics and social science as a whole. I get that question a lot less. I think that's growth in the field.
Rohan: Are there big questions about Indian politics that you want to see studied?
Pavithra: One is, I think we should constantly test our assumptions of how politics work. If you think about some of the most interesting recent works on Indian politics, Francesca Jensenius questioning "Have reservations really mattered?" Tariq Thachil asking, "Why are STs voting or in support of the BJP?" Or Milan Vaishnav asking, "Why do voters, knowing people are corrupt and criminals, still vote for them?" And more recently, someone like Feyaad Ali questioning the idea that there is such a thing as a uniform Muslim vote. All of these are going back to some first principles questions on, does politics work the way we think our inherited wisdom tells us it works? And if not, why is politics the way it is? I think we should keep doing that, go back to pushing the what-we-think-we-know questions.
But more broadly, I think India is at this funny crossroads of the North-South divide, which I'm intrigued by. And I want to understand more about the institutional, economic, and social dynamics of why that divide is so clearly emerging. This is all to say, I hope more people will study Tamil politics and Bengali politics and Malayali politics, because I think there's a lot to be learned from why politics look so different in one part of the country versus another.
Rohan: Are there recommendations for those interested in this space?
Pavithra: I have learned a great deal from American politics scholars within the course of my Ph.D. Ira Katznelson was the one who started getting me on this path of thinking about the relationship between identity and the stage. If people haven't heard of him or read his work, Ira Katznelson is a giant and his book, When Affirmative Action Was White, was really influential for me. And more recently, the US too is on a journey to understanding why social status is key to how we think about politics. Kathy Cramer’s The Politics of Resentment and Ashley Jardina’s White Identity Politics, I'd very strongly recommend.
Going back to what I was saying earlier, we need more scholarship questioning the first fundamental premise of things. So, I can't recommend highly enough Jennifer Bussell's book, Clients and Constituents, which pushes us to get away from this thinking that India's MLAs and MPs are not doing much, but instead they actually are doing a lot of constituency service. And what that looks like doesn't again conform to our preconceived notions of Indian politics. I think gender and gender's intersection with caste is a recently growing important part of research on Indian politics. Please read anything by Francesca Jensenius and more recently, Tanushree Goyal, Rachel Brulé, and Soledad Prillaman, who are pushing the envelope on thinking through why and how gender matters in Indian politics.
Pavithra Suryanarayan is an Assistant Professor in the Government Department at the London School of Economics and Political Science. In August 2024, she will be an Associate Professor in the Government Department at the LSE.
Rohan Venkat is the Consulting Editor for India in Transition and a CASI Spring 2024 Visiting Fellow.
As millions of Indians set out to vote over the next two months, India in Transition brings you CASI Election Conversations 2024, an interview series featuring renowned scholars reflecting on the factors and dimensions of politics, political economy, and democracy that will define India’s 2024 election. Earlier in the series, we featured Louise Tillin on federalism in India, Yamini Aiyar on the BJP’s “Techno-Patrimonial” welfare model and Rachel Brulé on the promises and pitfalls of gender quotas.
India in Transition (IiT) is published by the Center for the Advanced Study of India (CASI) of the University of Pennsylvania. All viewpoints, positions, and conclusions expressed in IiT are solely those of the author(s) and not specifically those of CASI.
© 2024 Center for the Advanced Study of India and the Trustees of the University of Pennsylvania. All rights reserved.