For several years, the caste census was presented by the INDIA alliance and its sympathetic commentariat as a moral X-ray machine that only Narendra Modi feared. The script was comfortable: the opposition asked brave questions about representation; the BJP allegedly hid behind ambiguity; television studios nodded in solemn agreement. Then the Union government decided to include caste enumeration in Census 2027, and the script developed an inconvenient second act. The Cabinet announcement explicitly placed caste enumeration inside the national Census rather than leaving it to a patchwork of state surveys. Suddenly, many people who had treated the demand as self-justifying had to confront the less glamorous questions of method, classification, privacy and use.
I welcome the decision, with one large warning attached. India cannot sensibly design targeted policy while relying on old, incomplete or politically curated pictures of social composition. But collecting data is not the same thing as accepting every claim made in its name. Rahul Gandhi’s slogan about population shares sounds arithmetically tidy because slogans do not have to reconcile overlapping deprivation, constitutional principles, regional differences or incentives. A Census can tell us who people say they are and where they live. It cannot by itself decide what justice requires. Anyone who promises that one giant table will settle India’s arguments is selling numerology with a social-justice label.
A national exercise beats partisan surveys
The government was right to put the work within the Census architecture. State-level exercises can produce useful information, but they vary in legal footing, questionnaire design, coverage and transparency. They can also become campaign instruments before statisticians have finished cleaning the data. A national operation at least offers common definitions, trained enumerators, comparable geographic coverage and established confidentiality rules. The official plan describes a two-phase process, digital collection and caste enumeration during the population phase. The government’s Census 2027 backgrounder also sets out the scale, technological changes and institutional preparations. That does not guarantee perfection; it does create a more credible starting point.
This is where the anti-Modi reflex becomes especially silly. Some critics first insisted that refusing a caste count proved bad faith. Once the government agreed, they shifted instantly to saying the decision was electoral opportunism. Politics influenced timing? I am devastated to discover politics in New Delhi. The relevant question is whether the chosen mechanism can produce lawful, usable data. If the answer is yes, a government is allowed to make a politically intelligent decision that also answers a legitimate policy need. The opposition does not own an idea forever merely because it shouted about it first.
The hard part is classification, not chanting
Enumerating caste is unlike asking whether a household has a tap. Names vary across states, spellings multiply, sub-castes may claim broader identities and respondents can use local terms unknown to a national coding list. Similar labels may describe different communities, while one community may be known by several names. Enumerators require careful training and a clear protocol that neither prompts answers nor humiliates respondents. The code directory must be tested publicly by specialists who understand linguistics, sociology and administrative categories. If this sounds tedious, good. Serious social policy is mostly tedious before it becomes a slogan.
The government should publish methodology notes, error ranges and documentation explaining how raw responses become categories. It should invite technical criticism without allowing activists or parties to rewrite the questionnaire midstream for advantage. Personal responses must remain confidential. Aggregated releases should be detailed enough for scholarship and policy, but safeguards against re-identification matter when small communities are mapped at fine geographic levels. Digital collection can speed validation, yet it also creates anxiety about centralised data. The answer is not theatrical panic; it is access control, auditing, minimal collection and clear penalties for misuse.
Counting is not a quota vending machine
My sharpest disagreement with the Congress-led discourse concerns what comes next. “Jitni abadi, utna haq” is a catchy line and an impoverished governing philosophy. Citizens possess rights as individuals; groups may require remedies for documented exclusion; public institutions also need competence, diversity and social legitimacy. These considerations cannot be collapsed into a single population ratio. If every public resource is allocated mechanically by caste share, democratic citizenship becomes a federation of hereditary claims. That is not emancipation from caste. It is caste furnished with a spreadsheet and a permanent secretariat.
The data should instead improve specific decisions. It can reveal whether benefits reach intended communities, whether educational gaps are narrowing, where occupational exclusion persists and how deprivation differs within broad official categories. It may show that some groups classed together have dramatically different outcomes. That finding would be administratively uncomfortable but valuable. Policy could then combine social disadvantage, household conditions, geography and educational indicators rather than treating each category as internally identical. The BJP, which has built broad coalitions among non-dominant communities, is better placed than its critics admit to make this case—but only if it resists turning the final tables into another exercise in electoral micro-targeting.
The opposition will face its own contradictions
Congress and several regional parties speak grandly about representation while operating organisations controlled by families, durable factions or socially dominant local networks. Once national data arrive, journalists should compare rhetoric with candidate selection, party posts and leadership succession. I suspect the moral certainty will become more elastic. Parties that demand proportionality everywhere may discover that applying it inside their own committees displaces familiar gatekeepers. This is why I have limited patience for the Lutyens habit of treating opposition slogans as inherently virtuous and government decisions as inherently suspect.
There is a fair opposition point the government should heed: delayed data are weakened data. India’s decennial Census disruption created a serious information gap. Welfare denominators, migration patterns and urban planning all suffer when administrators work from ageing population assumptions. The Centre should publish an achievable release calendar and explain any slippage promptly. It should not hide behind the magnitude of the exercise after choosing digital tools partly on the promise of faster processing. Credit for taking the decision must be matched by accountability for delivering results.
Prepare the public for inconvenient findings
Good data rarely flatter every political tribe. The Census may undermine claims repeated for decades. It may reveal unexpected mobility in one community and stubborn deprivation in another. It may show large interstate differences that make a single national slogan useless. The mature response will be to commission analysis, compare administrative datasets and adjust schemes carefully. The immature response will be to accuse the enumerators of conspiracy whenever a preferred estimate fails to appear. Unfortunately, our politics has industrial capacity in the latter.
The government should establish an expert advisory process whose membership and minutes are public. Parliament should debate principles for using aggregated caste data before the numbers trigger a bidding war. Courts will inevitably scrutinise policies influenced by the exercise, so legal reasoning cannot be added as an afterthought. State governments need guidance on comparability, and researchers need anonymised access under rigorous rules. These safeguards are not ways to dilute social justice. They are how we prevent social justice from becoming a licence for whichever party produces the loudest denominator.
I regard the caste census decision as a deserved political win for the Modi government because it called the opposition’s bluff without dismissing the underlying question. But a win in tactics is not yet a victory in governance. The measure of success will be a credible count, protected personal data and policies that address real disadvantage without freezing every Indian inside an inherited box.
The INDIA alliance wanted the question asked. It has been asked. Now it must offer answers more sophisticated than percentages shouted from a stage. The government, meanwhile, must show that its instinct for political jujitsu can be followed by administrative patience. I am glad India will have better information. I will be even happier if we refuse to let that information become the next excuse for reducing 1.4 billion complicated lives to a quota auction.
