On Counting Caste and Destroying It

In light of the recent debates about the caste census in India, I decided to read B.R Ambedkar’s Annihilation of Caste. Ambedkar’s bare, crisp, morally forthright case for the destruction of caste in India stands in interesting contrast to the more slippery discourse around whether the Indian state should collect data on caste membership as part of the next census. Interestingly, the debate does not follow the expected ideological fault lines and in the final count is at risk of clouding the goal of eradicating caste.

The current BJP led Union government, is either opposed, or at least ambivalent, about conducting a national census of caste membership. Despite being in power for a decade, it has also not stopped reservations or social programs to alleviate caste. This combined with the way it has trashed a more than 100 year tradition of a decennial census, means it aims to keep administering policy without data.

What becomes confusing is that the ideological foundry of the BJP, the RSS, have, at least publicly, come out in favour of the caste census. Initially in 2023, it was reported the RSS was against the caste census because it would harm the unity of the country and harm attempts to eradicate caste.¹ Others like Alok Bansal and Yashawardhana writing in the Indian Express have voiced similar concerns:

…adding caste to official Census data is bound to aggravate social differences, strengthen identity politics, and splinter the country along caste lines. Thus, this is by no stretch of imagination a harmless exercise. It could derail the movement towards the creation of a casteless society and has the potential to permanently damage the national integrity of the country.²

However, in September of 2024, the RSS announced it was in favour of the caste census for the purpose of “welfare” but not for it to be used as a “political tool”.³ In all this I think a certain level of cynicism about the RSS’s stance is warranted. Its claims from 2023 and 2024 appear to be more concerned with the effect that a caste census might have on their electioneering which is based on developing a unified Hindu national identity. After all, if a caste census is done it has the potential to show the voting strength of some castes over others, and if it is used by other parties it has the potential to break the BJP’s current cross-caste coalition of voters. To see how plausible this is one only has to look at the way the BJP has won the Uttar Pradesh election with hindu priest, Yogi Adityanath as its leader.

In all this the Congress and other opposition parties, find themselves on the same side as the RSS, calling for a caste census as a way to collect data on social discrimination. The case in favour of it is fairly straightforward. It is necessary to measure social discrimination, to fulfil certain constitutionally mandated goals of social justice, to administer government policy, and to stop certain castes from having an outsized share of the nation’s wealth. K. Ashok Vardhan Shetty articulates this well in the The Hindu, writing:

The problems of social groups that have been historically discriminated against (be it by caste, race, religion, gender, disability etc.) cannot be resolved without collecting data group-identity wise. Doing so is not a capitulation to identity politics but a vital step towards informed policy making and inclusive development.⁴

Shetty’s point is that counting caste membership is a way to measure the social discrimination that follows people identifying as a member of caste and acting according to the rules of the caste system. Social discrimination is the basis of that caste system. So, counting caste would throw light on the extent and nature of that discrimination which can then be redressed.

Shetty’s point is that counting caste membership is a way to measure the social discrimination that follows people identifying as a member of caste and acting according to the rules of the caste system. Social discrimination is the basis of that caste system. So, counting caste would throw light on the extent and nature of that discrimination which can then be redressed.

So what is the hesitancy with conducting a caste census? I suspect, although I do not have any evidence, there are two reasons. One is that some in the intellectual and political class, still hold the received religious view of caste, which is to see it as a metaphysical truth. Another is that they see caste as an unfortunate social construct that has deep roots in India. In both cases, a census would be an uncomfortable view of the social hierarchy.

To my mind, the correct response is not hesitancy, but to first acknowledge what ends a caste census would serve. Following Ambedkar, I think, Indian society should eradicate caste. To that end, what would be the point of counting something that one seeks to destroy? A caste census is not just a matter of counting an easily observable fact. It is a matter of asking people what caste they belong to. So it inherently involves an invitation for people to judge and classify themselves according to a social hierarchy we intend to destroy.

There is a clear parallel here to the case of “race”. To put it bluntly, ending racism, or racialism as it was once called, requires people to stop racialising themselves. One can still talk of ethnicities as descriptions of populations from a particular geography and history. But, to end discrimination on the 19th century idea of human beings as composed of distinct races with distinct destinies and immutable qualities, requires people to stop classifying themselves into a “race”.

How then does a society tackle caste discrimination without a census of caste membership? The question rests on a false premise. To know how much caste discrimination is going on requires counting instances of caste discrimination. That is what will tell a government how much caste discrimination exists, where it is happening, and who is involved in it.

But, you might say, what about caste discrimination or disadvantage that is the result of social norms, or economic development? Here the instances of caste discrimination are backed up by a system of social rules, rather than individual prejudices. How should a government gather information on how much to fund social programs in certain areas, or what level to set quotas or other forms of affirmative action? Program budgets and bureaucratic effort should be commensurate to the size of the castes that they seek to help. Achieving that requires a caste census doesn’t it?

In answering this question it is useful to consider Ambedkar’s discussion of social reform in Annihilation of Caste. For Ambedkar, social reform must precede political reform or economic reform because without destroying caste or at least dealing with the problems it creates, true political or economic reform is not possible. For Ambedkar, to reform hindu society in order to destroy caste is to “destroy the belief in the sanctity of the shastras”. This is what will allow people to have inter-caste marriages and socially interact as equals. For Ambedkar, this reform ought to involve a standardisation of hindu scripture to which all Hindus can accept. This will mean reconstructing Hindu doctrine on a “basis that will be in consonance with liberty, equality and fraternity; in short, with democracy”.⁵ The reform should also involve reestablishing a priesthood class which is open to all, appointed and regulated by the state.

Ambedkar argues that political reform is intimately connected to and must follow social reform. This is because since political reform in India rested on the idea that Indians ought to rule India, it raises the question of whether those Indians who are comfortable with the oppression of Indians by other Indians are worthy of political power. Moreover, reforms that redesign the political constitution of a society must take into account the divisions within society surrounding religion and social interactions. This is because in some sense these divisions must be dealt with to enable a peaceful society.

Against socialists who advocate for economic reform, Ambedkar argues that they see economic power as the only real power in society. But, Ambedkar argues, religion in India shows that the social status of an individual is a source of power and authority other than economic power. Moreover this social power or lack thereof is so strong it prevents any real economic reform or revolution because it does not allow any bonds of solidarity and fraternity to form between people. So, much like political reform, economic reform must wait for social reform.

Overall, I think Ambedkar overplays the priority of social reform over political and economic reform. Social reform to destroy the caste system, which in its most basic form is changing people’s beliefs about the social order, is a hard thing to do. Firstly, it is not clear that any religion’s doctrine has ever met Ambedkar’s criterion of being in consonance with democracy. The type of social reform Ambedkar seeks has no useful precedent to follow. But, even weaker reform has not been accomplished after 75 years of independent rule.

It is much easier to reform a religion when done together with political and economic reform. A social movement to “destroy the authority of the shastras” would require some political reform that recognised certain rights and liberties, formally banned caste discrimination, and created institutions capable of defending the reforms. Ambedkar believes that such political reform would be ineffective since it would be rejected by the community, and rights can only be protected by the “social and moral conscience of society”.⁶ This is not quite true. Of course a society’s fundamental laws must be acceptable to a society, but this acceptability is engendered not just from the bottom upwards. It is also engendered, at the same time from the top downwards by defending it robustly. Enforcing the law and punishing one as an example to others also changes society and its culture.

When it comes to economic reform, Ambedkar is right to point out that socialists would be wrong to say economic power is the only kind of power. It would also be wrong to downplay how economic reform can more easily break down a caste system. Economic reform, particularly in the form of urbanisation and industrialisation, are powerful tools for breaking down social hierarchies. The best example for this is Europe, where the industrial revolution and the rise of capitalism broke the old feudal system (a system that was functionally very similar to the Hindu caste system). But, this did not happen in India. For instance, in 1951, 68.1%⁷ of workers were employed in agriculture and 40 years later in 1991 that share had only lowered to 66.7%⁸. A socialism in India that was geared towards industrialisation and urbanising the population rapidly would have led people to voluntarily give up their belief in the authority of the shastras because they would have to work amongst people of different castes in formal organised employment and live next to them in dense cities. This sort of top-down structural change to social arrangements would have broken the hold of caste based social policing of small towns and villages. It would also have made it easier to spread caste blind educational and employment opportunities to the oppressed castes. All these economic changes would have aided a parallel social and religious reform.

In my view one major reason why caste still exists in Indian cities, which its remarkably dense urban centres, is because, as Ambedkar rightly points out, “The caste system is not merely a division of labourers—which is quite different from division of labour—it is a hierarchy in which the divisions of labourers are graded one above the other.”⁹ This rigid division of employment of the caste system was able to be transferred to the cities because of a lack of industrialisation. Consider, even in FY1999-2000, only 32.8% of urban male workers were employed in the secondary sector which comprises mining and quarrying, manufacturing, electricity, water, etc. and construction, while 60.6% were employed in the tertiary sector which comprises trade, hotel and restaurant, transport, storage and communication and other services. In FY2022-2023, the share of employment in the secondary sector has only risen to 35%.¹⁰ While living in cities afforded more educational and employment opportunities those opportunities were still prescribed by the caste system. Rapid industrialisation would have allowed Indians of the oppressed castes to find employment in manufacturing, construction, and services, without a college education.

Returning to the earlier question of whether a caste census is required to combat caste discrimination that is the result of social norms or economic development, the answer in my view for socialists, is that it is not required. Ambedkar is correct that destroying caste ought to be the primary goal if one wants to combat caste discrimination. Doing this requires an active social movement to end caste self-identification. It also requires social programs, quotas and other forms of affirmative action, to be based on income and wealth not caste membership because caste membership is what we seek to destroy. This is what is required to fulfil the socialist goal of providing economic equality so that all may enjoy the fruits of the economy. It would, importantly, from the perspective of the Indian welfare state, treat people with respect to their economic status as opposed to their membership of a caste. In all likelihood most recipients would in the short term be from the oppressed castes, there would not be much change in who receives the benefits. Importantly though, the explicit goal would be the uplifting of people sans a caste identity as opposed to uplifting caste groups. It would also have the added benefit of being universal welfare policies which in the long run tend to have stronger support.

I understand this might seem too hopeful and unrealistic. How can socialists be so sure changing economic circumstances will melt away the caste system? For instance, Ambedkar says against the socialist:

The assurance of a socialist leading the revolution that he does not believe in caste, I am sure, will not suffice. The assurance must be the assurance proceeding from a much deeper foundation—namely, the mental attitude of the compatriots towards one another in their spirit of personal equality and fraternity. Can it be said that the proletariat of India, poor as it is, recognises no distinctions except that of the rich and the poor? Can it be said that the poor in India recognise no such distinctions of caste or creed, high or low? If the fact is that they do, what unity of front can be expected from such a proletariat in its action against the rich? How can there be a revolution if the proletariat cannot present a united front?¹¹

Ambedkar is right to question whether any sort of socialism can be achieved if the poor and the oppressed themselves see caste as a relevant distinction. But, there are two responses that can answer the charge. The first is that we can put aside Ambedkar’s assumption that a revolution is the only path to socialism. It can be incremental and proceed from ordinary parliamentary politics. In fact one might think it has to be incremental since capitalism in industrialised economies did not immiserate the working class as predicted by socialists. The second response is that, if socialists need to inculcate, as Ambedkar says, a deeper foundation of equality and fraternity between people, it is highly unlikely a caste census will help this. Asking people to identify with a caste is not going to help people see others as fellow citizens entitled to the benefits of the industrialised economy and the welfare state.

Despite my critical reflections on the Annihilation of Caste, it offers a morally cleansing discussion of the social discrimination at the heart of the caste system, and the case for why and how it can be destroyed. To me this is its greatest value, since to my eyes at least, there is no widespread social reform or economic reform in India at the moment that seeks to dismantle caste with any urgency.