I have a confession that will appal people who think politics begins and ends with television debates: I enjoy reading census manuals. Budgets tell us what a government wants to do; a census tells us whether it knows whom it is governing. That is why the Registrar General’s decision to have states monitor Census 2027 population-enumeration work in real time strikes me as more consequential than most of this week’s noisy partisan quarrels.
The census is finally becoming a digital operation. The official Census of India technology overview describes this as the first fully digital Indian census, while a Home Ministry parliamentary reply records the two-phase design and the contemplated mix of apps, portals and offline alternatives. That is not glamorous reform. It is the plumbing of a capable state.
Measurement is infrastructure
The opposition habit is to treat administrative modernisation as either publicity or conspiracy. If a portal opens, somebody immediately warns of surveillance; if the government uses paper, the same voices complain that India is technologically backward. I find this reflex exhausting. A country of India’s scale cannot administer welfare, urban planning, school capacity, migration policy or constituency boundaries through decennial guesswork and spreadsheets passed between offices.
Real-time monitoring has a plain purpose. If one block reports suspiciously few households, supervisors can investigate while enumerators are still nearby. If an app version produces inconsistent answers, the error can be isolated before it contaminates a state-wide dataset. If a language field confuses respondents, the questionnaire can be clarified. In the paper era, many such problems became visible only during later compilation, when correcting them was expensive or impossible.
I also give the government credit for proceeding with a digital census after the long pandemic delay. Delay deserves criticism; pretending otherwise would be sycophancy. India should not have had to rely for so long on 2011 population structures while cities expanded, migration patterns changed and entire welfare architectures were rebuilt. Yet the sensible response to delay is to improve the eventual count, not to preserve an antiquated process as punishment.
The caste question raises the stakes
Census 2027 will matter politically because caste enumeration can rearrange arguments about representation and benefits. Public debate has relied too often on confident claims derived from old or partial data. The Congress and its allies like caste numbers when they imagine those numbers will unlock a convenient coalition; many of the same leaders become suddenly vague when asked how categories should be verified, reconciled or used.
I support collecting better data. I do not support pretending that collection automatically produces a policy. A census can describe social composition; it cannot decide the proper balance among poverty relief, historical disadvantage, regional inequality and individual opportunity. Those are political choices. Rahul Gandhi’s style of converting every dataset into a pre-written slogan may be electorally tempting, but a serious government must resist policy by hashtag.
The BJP should resist a temptation of its own: treating administrative control as sufficient evidence of public trust. Digital systems are not neutral merely because officials designed them. Software encodes definitions, validation rules and escalation procedures. A central dashboard can illuminate local failures, but it can also encourage local officers to make numbers look tidy for their superiors. Anyone who has worked around targets knows the danger: measurement improves conduct until the target becomes the conduct.
Publish the machinery, not personal data
The solution is not to abandon digitisation. It is to publish the operational safeguards around it. The government should release app-version histories, aggregate correction rates, district-level completion information, documented outage periods and the rules used to flag unusual submissions. None of this requires exposing household records. It would let statisticians, state governments and citizens distinguish quality control from arbitrary intervention.
Offline enumeration must also remain a genuine fallback. India contains poor-connectivity zones, elderly respondents uncomfortable with self-enumeration, migrant households with complicated residence patterns and citizens who reasonably want a human explanation before sharing information. “Digital first” should mean fewer transcription errors and faster supervision, not a bureaucratic excuse to mark difficult households as unreachable.
I would go further and commission an independent post-enumeration survey whose methodology and broad findings are public. Every census has undercount and duplication. A mature administration measures those errors rather than claiming perfection. Publishing an error estimate would not weaken the exercise. It would signal that the government understands statistics better than its critics understand political theatre.
Delimitation needs credible inputs
The shadow over the count is delimitation. Southern states worry that successful population stabilisation will reduce their parliamentary weight; faster-growing states argue that equal representation cannot be frozen indefinitely. Both concerns are legitimate. This dispute cannot be wished away with sentimental invocations of federalism, nor solved by Congress leaders promising incompatible things in different capitals.
The first requirement is a credible denominator. If political actors doubt the census, every later allocation or boundary decision will inherit that doubt. The government therefore has a self-interested reason to overinvest in transparency. It may have the constitutional authority to conduct the exercise, but authority is not the same as persuasion.
My own view is that representation and fiscal reward need not be collapsed into one crude formula. Parliament can recognise population change while protecting states from being punished financially for demographic success. But that debate should follow the count, not corrupt it. Enumerators must not become foot soldiers in an argument about seats.
A test of the Indian state
For decades, fashionable commentary treated state capacity as a dull concern compared with ideology. Then a pandemic, welfare transfers, internal migration and urban stress reminded everyone that an incapable state cannot deliver even a beautifully worded promise. The Modi governments have generally understood the importance of digital public infrastructure. Their weakness has sometimes been an impatience with explaining how systems work and how mistakes are remedied.
Census 2027 is an opportunity to combine ambition with institutional humility. Build the dashboard. Track anomalies quickly. Use encryption, access controls and accountable supervisors. But also show citizens the audit trail surrounding the machine. The government should not accept the opposition’s lazy claim that every digital reform is sinister. Nor should it hand that opposition avoidable ammunition through opacity.
I want the count to succeed because almost every serious policy argument depends on it. Credit is due for dragging the census into the digital age. The next reform is to let the public see, in aggregate, how that digital age corrects itself. A government confident in its administrative achievement should be confident enough to publish its error log.
