Researched by: Sara Varghese 

Some surprising background

Here are some facts you may not be aware of when it comes to castes and numbers in India.

Fact #1: We have not counted the number of castes or the people who belong to each of them in independent India. The Indian Census under British rule included all castes, but the first census in free India only counted members of scheduled castes and tribes. Everyone else is thrown under the ‘general’ category. And that has remained the norm ever since. 

Fact #2: This means we do not know the caste of three-fourths of all Indians. We don’t even have an official list of all castes in India. 

Fact #3: We also have no clue about the number of people who fall under the Other Backward Classes (OBC) category—the vast swathe of lower and intermediate castes that lie between the upper castes and Dalits. Therefore, our affirmative action schemes for these castes are arbitrary—and have zero basis on actual demographics. 

Example: The Mandal Commission in the 1990s estimated OBCs as comprising 52% of the population—based on 1931 British-era data! And the government then randomly assigned a 27% quota to OBCs.

Fact #4: In 1992, the Supreme Court placed a cap of 50% on all reservations—which was not based on any demographic data. And in 1993, it ordered the government to ensure that the “creamy layer” of any backward caste be excluded from reservations—since “seats and posts reserved for backward classes are snatched away by the more fortunate among them.” Again, since there is no socio-economic data for OBCs, this ruling too was made in the absence of any hard data backing up its claim.

Fact #5: In 2011, an under-pressure UPA government reluctantly conducted the Socio-Economic Caste Census (SECC 2011). But the details were never released because it was apparently riddled with “crores of errors.” Also: “The number of castes ran into lakhs.” OTOH, the unpublished data has been used by government schemes to identify beneficiaries ever since. 

Fact #6: There are however regular national surveys—based on samples not actual counts—that offer some indication of caste in India. For example: the National Family and Health Survey (NFHS) and National Sample Surveys (NSS). And these typically reveal what we already suspect:

“It is seen that a majority proportion of ‘upper’ caste households report their principal occupation in public services/white-collar jobs category, while the majority of STs in agriculture/fisheries, SCs in manual labour, and OBCs in blue-collar jobs. It is ironic that even after years of Independence, the occupational profile of the population still coincides, to a large extent, with the unjust division of labour as prescribed in the archaic Varna system.”

Why should we count caste?

The opposition parties are unified in demanding a caste census—as are anti-caste activists. Also pushing for one: BJP allies like Janata Dal (U) in Bihar. And here’s why.

One: Caste is the single most important factor that determines a person’s quality of life in India. The argument goes something like this: Caste is real; and we must document that reality. Also, documenting caste does not mean legitimising it, as Tejas Harad explains: 

“There is this misconception amongst the upper castes that if we talk about caste, it will perpetuate caste. And that is there right from the beginning. If you look at Congress leadership—right from Jawahar Lal Nehru—their approach to this whole problem was: ‘if we ignore it, it will, somehow, because of modernity, it will get destroyed or somehow pass away...But that is a complete lie. If we have to solve a problem or find a solution, we have to acknowledge that a problem exists and in what condition the problem exists and that we can understand only if we conduct this caste census.”

Two: Forget the number of people who belong to each caste, there is also zero socio-economic information about them. For example: education, infant mortality rate, child marriage etc. What we have instead is a constant clamour of different castes to be counted as OBCs—to jostle for benefits in a reservation-based system. And then there is talk about that ‘creamy’ (privileged) layer of these castes—who apparently do not deserve government support. Documenting the socio-economic status of these castes will help bring hard facts to these heated political debates. As Yogendra Yadav notes:

“Such a data-set would put to rest street agitations about inclusion within OBC lists. So, if Jats or Marathas want to be included in the OBC list, they would just dip into census data and prove their ‘backwardness’. Or else, they would have to keep quiet. Over the years, this is precisely what the courts have demanded: Any reservation must be based on verifiable empirical evidence. What better evidence than the census?”

Three: Our affirmative action programs based on caste are among the biggest in the world—and yet most are not based on any kind of hard data. And we are not tracking the socio-economic condition of OBCs covered by these programs:

“To this date, the outdated data of the 1931 census largely remains the basis for reservation in the country. Where deprivation is associated with historical exclusion from access to education or valuable resources rather than active social discrimination, it becomes important to take stock of the changing situation of such communities at periodic intervals. Therefore, obtaining accurate data for better targeting of reservation policy becomes even more essential.”

Four: Others argue that a caste census may reveal uncomfortable facts that we will be forced to grapple with. For one, the upper castes may constitute an even smaller minority than they fear—and a census may make their disproportionate power glaringly obvious. For example, the Economic Times dryly noted: 

“According to a story making the rounds, when the caste data [of the 2011 Socio-Economic Caste Census] was compiled a top official was so startled to see the upper caste numbers… that he immediately jumped into his vehicle and sped towards Raisina Hill to share the findings with his bosses. They too were convinced that the upper caste numbers are dangerously low to be revealed to the world.”

Point to note: In 2014, Congress-ruled Karnataka ordered a caste-based survey to rationalise its OBC reservation quotas. But then it promptly suppressed the results—because they showed that Scheduled Castes were the largest community, followed by Muslims. And that posed a problem for the powerful Lingayats and Vokkaligas—who are assumed to be the more numerically dominant—as it would upset traditional vote calculations. As one insider said: 

“The opposition came not from a community perspective; they do not worry that their community will lose out on benefits. They don’t worry that their community people will not get education or jobs. They think that if the report is made public then the Lingayats and Vokkaliga leaders will no longer be able to hold important and influential positions.”

Five: A caste census also serves the interests of regional parties—who often owe their political power to caste-based appeals. This is even more pronounced in states like Bihar and UP where caste calculations determine the outcome of elections. For example: If the percentage of OBCs is above 52%, these parties will campaign for higher quotas—making that the centrepiece of their poll platform. 

About those naysayers…

One: A number of political leaders believe that a caste census will unleash total mayhem:

“This might lead to an all-out caste war. The castes whose numbers get reduced will surely question the process involved in such an exercise. Their quotas might also be truncated. On the other hand, castes emerging with greater numbers will hanker for a bigger share of the reservation pie. Fresh demands will be raised to raise the current ceiling on the reservation quota. Again, it will not be an easy process to accommodate all the castes in that increased reservation quota. No government would want such a situation.”

Two: The classic Nehruvian argument is that counting caste will reaffirm its power—and undo efforts to eradicate it. The supporters of that argument includes the RSS:

“On May 24, 2010, when the debate on the subject had peaked ahead of Census 2011, then RSS sar-karyawah Suresh Bhaiyaji Joshi had said in a statement from Nagpur: 'We are not against registering categories, but we oppose registering castes.' He had said a caste-based census is against the idea of a casteless society envisaged by leaders like Babasaheb Ambedkar in the Constitution and will weaken ongoing efforts to create social harmony.”

Three: While the BJP has done very well among OBCs in the recent national election, its success requires the consolidation of a shared ‘Hindu’ identity, as BBC News notes:

“Caste is a crucial factor in every Indian election, from the village council to the parliament. More so in Uttar Pradesh, where the BJP's power and popularity rest on a delicately forged alliance of castes, and especially those in the OBC category. A caste count could cause fissures in the Hindu vote, which the BJP has managed to consolidate in recent years, despite deep divisions that underpin the party's plank of Hindu unity.”

Also: counting OBCs may offer an uncomfortable data point for a party that remains dominated by upper caste leadership and values.  

The bottomline: Who we count depends entirely on who counts in a deeply hierarchical society like India.

Reading list

• BBC News offers an excellent overview of the history of counting caste in India—and the politics involved. 
• Yogendra Yadav in The Print, Shoaib Daniyal in Scroll and Mayank Labh in Leaflet offer strong arguments in favour of a caste census. 
• Indian Express explains why the BJP isn’t keen on counting caste.
• Rakesh Chandra in The Print argues that we need more than just a head count when it comes to documenting caste.
• Hindustan Times decodes the political motives surrounding the caste census.
• Economic Times excellently skewers upper caste fears about a caste census.
• And The News Minute has the inside story on why the Karnataka government killed its census data.