For years, the fashionable Delhi line was that the BJP feared a caste census because accurate numbers would disturb its coalition. Then the government decided to include caste enumeration in Census 2027, and the same people pivoted without pausing for breath: now the decision was belated, tactical or somehow stolen from the opposition. Lutyens commentary has a marvellous immunity to falsification. Whatever Narendra Modi does confirms the theory announced before he did the opposite. I see the decision more plainly. India cannot sensibly target welfare, debate representation or understand social mobility using a mixture of 1930s-era assumptions, partial state surveys and political claims manufactured at rallies. The Cabinet announcement placed caste enumeration within the national Census rather than commissioning another freestanding exercise. That institutional choice matters. Rahul Gandhi may have made “jitni abadi, utna haq” a slogan, but a republic needs definitions, procedures and privacy safeguards, not a microphone followed by arithmetic.
Counting is not the same as surrendering to caste politics
Some people on the Right regard any caste count as capitulation. I understand the anxiety. Caste statistics can be weaponised by parties whose entire business model is slicing citizens into bargaining blocs. Numbers acquire a false moral certainty: if a group forms a given share of the population, activists leap to the conclusion that every job, seat, budget and institution must mirror that share. That is not social justice; it is a permanent census auction. Yet refusing to count does not dissolve caste. It merely leaves the field to estimates selectively quoted by politicians. Bad data does not produce casteless citizenship. It produces louder brokers. In my view, a credible national enumeration can actually weaken their monopoly, provided the government separates descriptive evidence from automatic entitlement. We should know how communities live, learn and work. We are not obliged to convert every table into a constitutional demand by lunchtime.
The official design offers reasons for cautious confidence. A government backgrounder describes Census 2027 as India’s first digital enumeration, with mobile collection, self-enumeration and a substantial administrative framework. Digital tools can improve validation and speed, though anyone who has watched a government dropdown menu encounter Indian social reality should remain humble. Caste names vary by spelling, region, language, sub-group and local usage. The same label can refer to different social locations across states; distinct names can refer to related groups. An enumerator cannot resolve generations of taxonomy during a doorstep conversation. The exercise therefore needs a publicly explained coding protocol, careful training, a process for uncertain entries and an expert audit that is neither a conclave of ideological favourites nor a secret bureaucratic chamber.
Congress wants the headline without owning the hard questions
The Congress posture is especially irritating because it treats intention as implementation. Its leaders demand a count, promise sweeping consequences and then behave as if administrative difficulties are excuses invented by hostile officials. But the party governed India for decades after Independence without conducting a full caste enumeration beyond Scheduled Castes and Scheduled Tribes. The government’s own earlier parliamentary account notes that post-Independence censuses did not enumerate other castes. Congress can argue that circumstances changed; fair enough. What it cannot plausibly do is present itself as the impatient victim of a prohibition imposed by the BJP. The failed handling of the Socio-Economic and Caste Census data should have taught humility about collection, classification and publication. Instead we get moral theatre: ask anyone for methodological detail and he is accused of fearing the truth.
The slogan “rights according to population” also collapses several different questions. Political representation, public employment, university admissions, poverty programmes and private economic opportunity are not interchangeable pots. Reservations respond to historical and social disadvantage within a constitutional framework, not solely headcount. Welfare may be better targeted by deprivation indicators that cross caste boundaries. Electoral representation is mediated by geography and democratic competition. Rahul Gandhi’s formula sounds radical because it declines to specify which right, against whose claim and under what legal rule. The media repeats it because ambiguity fits a chyron. A responsible government must be duller. It should publish the data with documentation, invite scholarship and evaluate policy domain by domain. The purpose of evidence is to improve judgment, not abolish it.
Privacy and trust cannot be afterthoughts
A digital census containing caste data will be among the most sensitive datasets the Indian state has ever assembled. I support the exercise, which is precisely why I refuse to wave away privacy concerns as elite fussiness. Household-level information must not become a searchable political directory. Access controls, encryption, retention rules and penalties for unauthorised use should be stated before collection begins. Published tables need adequate aggregation so that small localities do not expose families. The government should explain whether raw records can be linked with other databases and, if so, under what statutory authority. Consent in a compulsory census is complicated; protection must therefore rest on law and architecture, not a line in an app promising security.
Trust also depends on publication. If the state collects caste data and then delays release because early tables become politically awkward, it will validate every suspicion. A staged calendar should be announced: preliminary population totals, cleaned caste tables, socioeconomic cross-tabulations and methodological reports. Revisions should be logged rather than quietly substituted. Independent statisticians should be able to reproduce aggregate findings using controlled access. None of this requires surrendering data sovereignty or inviting activists to rummage through identifiable records. It requires recognising that credibility is built when outsiders can understand how a conclusion emerged. The government has the administrative muscle to conduct the count; it must now show the confidence to make its workings inspectable.
What the numbers should change
If done well, the census can improve policy in ways less dramatic than quota politics and more useful to citizens. It can reveal communities with persistent gaps in schooling, housing, disability support or workforce participation. It can show whether broad categories conceal groups that have advanced and others that remain excluded. It can help evaluate whether schemes reach intended beneficiaries without assuming that every member of a caste shares one economic condition. I would combine caste tables with district-level indicators and programme outcomes, while protecting individual records. Policy should then favour measurable disadvantage and periodically test whether interventions work. That is more demanding than distributing symbolic promises, which is exactly why professional agitators prefer slogans.
The BJP should resist two temptations. The first is claiming that the decision settles the political argument. It begins a harder administrative one. The second is using Congress’s hypocrisy as an excuse to dismiss all opposition scrutiny. Questions about categories, field testing and privacy are legitimate even when raised by people I usually distrust. A confident government can answer them without adopting the commentariat’s framing. It can also remind citizens that social cohesion matters: data should illuminate inequality without teaching every child that ancestry is destiny.
I support caste enumeration because governing a complicated country through folklore is irresponsible. I support placing it in the Census because national standards beat a patchwork of politically commissioned surveys. And I support the government’s measured approach over Congress’s instant entitlement calculus. But support is not a blank cheque. The count must be technically credible, safely held, promptly published and interpreted with constitutional sobriety. India needs the truth about caste; it does not need another industry devoted to shouting percentages at one another. If Census 2027 produces knowledge rather than ammunition, it will be a reform. If parties reduce it to electoral inventory, the enumerators will have done historic work only for politicians to make it small.
