Interview International Video Interviews

Castes and State in India
Interview with Rohini Somanathan

by Jules Naudet , 19 April
with the support of CASBS

Quotas in India contribute to the emancipation of lower castes while producing perverse effects that are difficult to control. Rohini Somanathan questions the right balance between targeted positive discrimination policies and public policies with a universal vocation.

This publication is part of our partnership with the Center for Advanced Study in the Behavioral Sciences. The full list of our joint publications is available here.
Rohini Somanathan is Professor of Economics at the Delhi School of Economics. Her research focuses on how social institutions interact with public policies to shape patterns of economic and social inequality. Within the broad area of development economics, she has worked on group identity and public goods, access to microfinance, child nutrition programs and environmental health. She is an elected fellow of the Econometric Society and the International Economic Association and a former 2022-2023 CASBS fellow at Stanford University.

Books & ideas: Your work explores the way the State shapes caste in India today. Before discussing this issue in the contemporary period, could you please remind us what scholarship tells us about the way caste emerged as an administrative category in the colonial period?

Rohini Somanathan : Caste really didn’t matter to the East India Company, which controlled India till 1857. So the first half of the 19th century, in the last quarter of the 18th century, and the only time that caste appears in documents of the East India Company is when there are concerns about rebellion or trade or revenue collection that are related to caste.

After 1857, things change dramatically with direct British rule. This is about the time that the censuses are getting rolled out in England and Wales, and the Indian census also begins in a limited way. And apart from asking the usual demographic information, they also ask about caste. And that’s when it really becomes an administrative category. And different censuses then record castes with different degrees of detail.

Some of them spend many, many chapters on caste. Some of them are more brief. And this really depends on how colonial administrators view caste and how important they think it is.

Books and Ideas: To what extent is caste an administrative category as much as a religious one?

Rohini Somanathan : It’s both administrative and religious and social and cultural. Once it gets recorded and encoded and published, it becomes administrative; and then it becomes even more so when the eligibility for employment or admissions to colleges start to depend on caste. So it’s all of those things. It’s a political category, it’s an administrative category, it’s a religious category, but in different senses of the word.

It’s religious to the extent that caste matters in ritual, it’s social, to the extent that castes have different occupations and they live in different parts of villages, often, and they might have rules of eating together or not eating together. It’s administrative in a different sense, because if a number of castes are clumped together in a category for affirmative action or for political mobilization, then that category is different in composition from the ritual category. So it’s all of those things, but in slightly different guises.

Books & Ideas: In what way did the Constitution of the Independent Republic of India mark a major rupture in the way caste is experienced daily?

Rohini Somanathan : The answer to this is complicated because on the one hand, the Constitution acknowledged equal citizenship and created clauses that would promote equal citizenship. On the other hand, it acknowledged the fact that people had differential access to resources in society, and it wanted to correct for that. So it also had clauses which encourage the state to take measures that would promote the success of disadvantaged castes and would also reduce the stigma associated with being lower caste or being what was called an untouchable caste.

And these things were happening simultaneously. So on the one hand, it really decried caste. But on the other hand, it encouraged to the including of caste. And so both of these things were working in in different directions. And so it wasn’t a major rupture. I’d say it was a shift. What it did do is bring caste into political and public life because it mandated reservations, what we call reservations.

And these were quotas for different castes and legislative bodies. And when it did this, it ensured there was more representation, but it also ensured that people had to think about caste when they would, deciding whether to stand for office. It determined access to political and public life. So it changed the way that caste was viewed in India. It was no longer something that was outside political life. It was very central to political life.

Books and Ideas: Since the 1990s and the publication of the Mandal report, caste has increasingly become a controversial and polarizing issue. How do you account for this increasing centrality of caste in political debates?

Rohini Somanathan : The Mandal report was the 1980 report of what’s called the Second Backward Classes Commission. The first Backward Classes Commission was in the 1950s, in 1955. And what actually happened was that this commission reviewed the categorization of caste for the purposes of affirmative action. So the Constitution in 1950 created two schedules. One was the list of scheduled castes, and another was the scheduled tribes. And what happened after this? Just because there are so many castes and such a diversity of names, there was some discussion of whether they did this right, whether people who were really disadvantaged appeared on the lists, whether they should have or not. And so there was a commission set up, and this was in 1955 to reevaluate this.

And they visited multiple states. They had meetings with people to really look at the whole categorization issue. And they listed 2400 castes as backward castes that were not on the original schedule. And so these were then called the other backward classes. And what the Mandal Commission did was review this again, because, again, there was disagreement. And part of the problem was really that the stakes were high.

So if there was some government category that was created and it didn’t count for anything, I don’t think there would have been the conflict that actually subsequently occurred. It counted for a lot because the most elite employment positions, the most elite universities had quotas for caste. And this meant that if you were part of a schedule, then it would influence your probability of getting these seats.

And so what the Mandal Commission report did was to suggest that there were many, many groups that were listed as the other backward classes, and they should also get reservations in government jobs and in universities. And because the public university system was still very limited, it hadn’t expanded, it was still hard to get into the best colleges. The stakes were very high.

And so what happened then is that the probability of some people increased based on their social origins and the probabilities of others decreased. And that was what led to the conflict.

Books and Ideas: What do we know about the ways in which caste scheduling shaped social and economic inequalities in India?

Rohini Somanathan : So scheduling changed access and it changed access largely to the elites sector. So scheduling really didn’t affect primary schools, it didn’t affect middle schools. It really mattered when you were trying to enter university or you were trying to apply for a government job. And so scheduling affected inequalities by increasing the probability of those that were eligible and scheduled for university and for public employment.

So what it did as a result was to increase inequalities within the scheduled groups. So groups that were with very limited access to primary education had no chance of entering these positions. And so inequalities between those groups, such as the Musahars and groups such as the Mahars and the Chamars who were much more educated, increased ; we know very little about what happened to those outside the scheduled groups because the census didn’t record information about them.

We only know what happened within the scheduled groups and what we can see, both for the caste and the tribes, was that within these groups, those that were more educated in 1961 were much more educated in 2011. Whereas those that were at the bottom of the ladder stayed at the bottom of the ladder. But they didn’t only stay at the bottom of the ladder. It was not only in terms of their rank, it was also in terms of the absolute changes that they experienced. They were relatively marginal.

And so we know this. We know that scheduling did bring some groups at par with the non-scheduled population. So when you have to think about what happens to inequality, you really need to start thinking about what data is available.

And here you have detailed data on the groups that are scheduled, and you really don’t have data on the groups that are not scheduled. So those are not broken up into different castes. They are just one big mass and you have the non-scheduled population and then you have all this data on the scheduled population. And so what you know from these data are that those inequalities increased some of the scheduled groups caught up with the non scheduled groups, but you can’t break up the non scheduled groups so you don’t know what happened there.

So that’s just in terms of the facts, it’s not saying why this happened. The conjecture is that you can only take advantage of scheduling if you already have some advantages and if you are really disadvantaged and marginalized, then you can’t. What’s also interesting in this context is that, even in 2011, which is the last census for which data is available, only 18% of Indian villages have secondary schools.

And so some of this is going to be determined by eligibility, some of it is going to be determined by access. And so we find that within the scheduled groups, the more advantaged gained more. And this is not to say that these were absolute advantages. There were still big gaps between them and a lot of the population. We just don’t measure those gaps very well. What we can do is look within groups that are scheduled.

Books & Ideas: Drawing upon your comparative research, what are the most promising public policy paths to follow if we truly want to combat social hierarchies of domination across the globe?

Rohini Somanathan : That’s a big, big question. And I think what I have been trying to do is precisely that, and I’ve been trying to look at it in the Indian case just because I know India best. It’s really interesting and it’s complicated. And I really believe there was a serious effort more than anywhere else in the world at the time of really tackling these inequalities.

So this is where I put my economist’s hat on. And I think that all public policies that try to reduce inequalities between groups, especially in these deeply divided societies, are of two kinds. One set is policies that affect selection. They’re not changing capacities, but they really changing the probabilities of entry into elite sectors. And affirmative action is in that class; anything that changes the probabilities of being selected, that is one set of policies.

The other set consists of policies that are on paper universal. So public schooling, health care, shelter, universal, basic income, food subsidies for the poor. All of these are policies that don’t have identity encoded into them, but they’re going to level things up simply because they’re changing capacities on the ground. And affirmative action and quotas come in the first set; all the basic needs policies come in the second set.

And the really big question that I’ve been I’ve been battling with is how do these two kinds of policies interact? Now, when the Constitution was written, the Constitution makers saw these complementarities. They saw these two kinds of policies as very complementary. And the reason is simple, straightforward. It makes sense: when you have political representation, then you ensure that your groups are represented on the ground.

And so if villages with concentrations of minorities don’t have goods, you bring that into the political arena. You bring public attention to draw on these issues, and it makes sure that access is more universal. So that’s the role that political representation plays. In addition, affirmative action creates role models. If you’re building schools, it’s going to be a very long time before you have groups that rise up through the school system and can be really competitive. They also don’t have access to the same networks that generations of very advantaged populations have.

And so what affirmative action does is create role models. You think to yourself, look, you know, if they look like me, if they’re from my village and they can do it, then so can I. And so the idea of Bhimrao Ambedkar, who was talked about as the draftsman of the Constitution, but also a lot of the others who were involved, a lot of the constituent assembly debates were about these complementarities, about how you need both identity based policies and universal public goods to work together. And the idea then was that you would only need the identity based policies for a while until universal public goods took over. And the playing field, as philosophers would call it, was really level. What happens in a democracy is that once you create political rents, then it has its own logic.

And as it turned out that the identity based policies, became more and more important, and the public goods, to my mind, were neglected. So if most populations in rural Bihar don’t have access to a secondary school or in rural Uttar Pradesh, then one has to ask, look, is this partly a result of the focus being shifted to identity based policies and away from universal policies?

And you want to ask the counterfactual: What if we had done otherwise? What if we really ensured that there was a school within a mile or two miles or three miles of every child? There were norms that were set down. What if those norms were really followed? And these kinds of questions are always hard to answer because you never observe what could have happened.

You have to just have conjectures about this. But what is interesting is that some of the scheduled castes that have been important in politics in North India have actually done better in South India, where they’re are less powerful, like in Tamil-Nadu, for example. And I think the reason they’ve done better is because southern India has really done much better on the public goods front.

So if you look at access in large villages of Tamil -Nadu, of Andhra Pradesh, of Kerala, Karnataka, access is much, much higher than it is in North India. And so my own conjecture is that we would have been better off to lay more emphasis on universal goods, and I think we should do that now. I can’t say that with complete conviction.

We can’t observe what if, but just looking at the history of inequality across the world, public education has been a big equalizer. In Northern Europe, many countries inherited a few inherit feudal systems. I think a little bit of what’s happened in India is that the politics has it’s broadened that a lot more caste represented. Now, many social origins are represented.

But it’s still about rent rather than raising everyone to a certain level. And I think that should be the focus in the years ahead.

Dossier's Articles

by Jules Naudet, 19 April

To quote this article :

Jules Naudet, « Castes and State in India. Interview with Rohini Somanathan », Books and Ideas , 19 April 2024. ISSN : 2105-3030. URL :

Nota Bene:

If you want to discuss this essay further, you can send a proposal to the editorial team (redaction at We will get back to you as soon as possible.

Our partners

© - Any replication forbidden without the explicit consent of the editors. - Mentions légales - webdesign : Abel Poucet