Census 2027: Counting India’s Missing Numbers

In the forested districts of Jharkhand—Palamu, Chatra, Khunti, Ranchi—where the soil runs red and jobs run scarce, hunger has a bureaucratic origin. Families have eaten wild herbs and survived on a single meal a day, not only because of poverty, but because the government did not count them.

Without an updated census, millions were excluded from the Public Distribution System, the network of subsidised grain that functions as India’s last nutritional safety net. The 2011 Census —the last to be completed—determined how many ration cards each state received under the National Food Security Act. By law, coverage can extend to 75 per cent of rural households and 50 per cent of urban ones. But the lists were frozen. Deaths allowed removals; new children and families did not earn automatic entry. The numbers stacked up in silence.

“Without data from the census, how can one make effective policy? Not doing the census is causing great deprivation for the poor and marginalised who need subsidised food to survive with dignity,” says Anjali Bharadwaj, transparency activist, Satark Nagrik Sangthan.

Now, 16 years after that last count, India has finally begun to count again. Census 2027—the country’s first fully digital enumeration, spanning 28 states, eight union territories, over 7,000 towns, and 640,000 villages—is underway. For rights advocates who spent years watching welfare allocations drift further from reality, the exercise is both overdue and potentially transformative. Yet a quiet comparison is already being made: not between Census 2027 and its predecessor, but between the national exercise and a more audacious one carried out by a single southern state.

The Circular Impasse

The 2021 Census was postponed on grounds of the COVID-19 pandemic—even as political rallies were permitted and state elections held across the country. For rights advocates, the justification never held. The consequence was a governance vacuum: welfare allocations, parliamentary delimitation, and infrastructure planning, all proceeded on decade-old snapshots of a country that had since added tens of millions of people and endured a historic public health crisis.

The Supreme Court, recognising the exclusion of migrant and unorganised workers, directed the government to issue 80 million additional ration cards as an interim measure. The government declined, citing the need for updated census data first—data it had itself failed to collect. It was a circular impasse with real human cost: the state withheld the remedy while withholding the precondition for the remedy.

The Bloodbath For The Poorest

On the ground, the consequences were not abstract. Families already enrolled in welfare programmes could not add newborn children—the lists were administratively closed. Meanwhile, the criteria for inclusion became quietly arbitrary.

In Rajasthan, automatic inclusions for the elderly and disabled were ignored because district quotas were full. Approximately four lakh disabled persons and nine lakh elderly residents in the state alone were left uncovered, according to the Mazdoor Kisan Shakti Sangathan (MKSS), the grassroots organisation co-founded by social activist Nikhil Dey.

“The more powerful get in via clout and appeals. Automatic inclusions like disabled and elderly pensioners are missed. On the ground, it’s a bloodbath for the poorest,” says Dey.

Digital exclusion compounded the problem. Household surveys used proxy markers of wealth—ownership of a four-wheeler, for instance—to remove people from eligibility lists. Gig economy workers or daily-wage drivers, who had access to a borrowed vehicle, could find themselves struck off.

At one point, any household with a two-wheeler faced exclusion —a criterion that would have disqualified much of the country’s working poor, who rely on motorbikes because public transport does not reach them. A targeting system designed to reach the neediest was, in practice, targeting the wrong people.

Enter Telangana: A State That Decided To Count For Itself

While the national government delayed, one state acted. India’s youngest state, Telangana—carved from undivided Andhra Pradesh in 2014—inherited data chaos alongside its new borders. With no credible baseline of its own population’s composition, the state launched what would become the most comprehensive sub-national caste and socio-economic survey India had seen in the post-independence era.

Dey, who served on the expert committee advising Telangana’s data exercise, describes it as a benchmark for what the national census could aspire to. Where Census 2027 asks approximately 30 questions, Telangana’s Socio-Economic Caste Census asked 85 questions with 285 sub-questions. Where the national form captures household assets as a blunt proxy for economic status, Telangana mapped income, profession, discrimination experiences, access to education and health, and the prevalence of child marriage. Where the national census lists caste as a column, Telangana built a framework for village-level comparison: how does the condition of a Dalit family in Nalgonda compare to the state average? How does a tribal household in Adilabad sit relative to its OBC neighbours? These are the questions effective welfare policy needs answered.

“Telangana’s socio-economic caste census adds many questions beyond just a caste column, enabling rich policy analysis. The SC/ST marginalisation was found to be three times worse than the state average. That’s the kind of finding that moves policy—and people,” says Dey.

The findings, released quietly just days before a parliamentary debate on the national census—with a proper press conference still pending, Dey noted—were striking. Scheduled Caste households showed deprivation indicators roughly three times worse than the state average. Among Scheduled Tribes, the picture approached near-total vulnerability: 94 per cent of SC households and 98 per cent of ST households were found to be economically vulnerable by the survey’s composite index. These were not numbers a 30-question form would have surfaced. The release itself was telling: quietly timed, poorly publicised, and only beginning to receive the scrutiny it warranted.

Census 2027 Vs. Telangana SECC: A Comparative Analysis

A side-by-side examination of Census 2027 and Telangana's Socio-Economic Caste Census (SECC) reveals not merely a difference in scale, but a fundamental divergence in governance philosophy—in what each exercise believes the state owes its citizens in the act of knowing them. Telangana’s SECC produced a depth of socio-economic profiling that the national exercise does not match.

While both exercises record caste identity, the national census treats it as a demographic column, whereas Telangana viewed it through an analytical lens—cross-referencing caste affiliation with income levels, access to education and healthcare, professional status, incidence of child marriage, and experiences of discrimination, enabling village-level comparisons of how marginalised communities fare relative to the state average.

On economic profiling, Census 2027 relies on household asset ownership as a proxy for financial status, while Telangana directly mapped income, occupation type, and access to essential services. Women's labour—including unpaid domestic work and wage differentials—is only partially captured in the national census, whereas Telangana’s framework treated it as a dedicated domain of enquiry.

The two exercises also differ on transparency: Telangana’s expert committee recommended that all data be placed in the public domain for civil society and research use, while Census 2027’s release model remains government-controlled. On methodology, both are digitally collected, though Telangana’s fieldwork was preceded by paper-based preparatory surveys.

As for findings, Telangana’s data has already surfaced that Scheduled Caste and Scheduled Tribe households face deprivation indicators roughly three times worse than the state average—findings of the kind that move policy. Census 2027’s comparable data remains to be released. Together, the two exercises mark a significant advance over 16 years of enumeration absence; but the Telangana model, in the estimation of those who helped design it, demonstrates what the national census could achieve if depth were treated not as a luxury but as a democratic obligation.

A Century Reconsidered: The Return of Caste Data

Caste was last counted comprehensively in the national census of 1931, under British administration. Its exclusion from every post-independence census was a deliberate decision rooted in Nehruvian anxieties about entrenching identity—a logic that, decades later, left policymakers unable to say with any precision how many OBC citizens India had, where they lived, or what they lacked. The Mandal Commission of 1980 worked from projections. Court judgments on reservation quotas relied on estimates. The world’s largest democracy was allocating life opportunities on the basis of demographic guesses.

Census 2027 breaks that silence. For Dey and Bharadwaj alike, the decision to include caste data is unambiguously welcome. But the lesson from Hyderabad is that a caste column alone is not enough. “Politics favours the powerful”, Dey observes, “but data arms the poor and marginalised to fight for resources”. The difference lies in depth: whether the data can answer not just how many, but how much worse—and why. Telangana’s exercise asked questions about discrimination, access to justice, and intra-community wage gaps. The national census, as currently framed, does not.

What Updated Numbers Could Mean

Advocates estimate that when 2027 data is published, between 120 and 150 million additional Indians—primarily migrants, the urban poor, and unorganised workers—should qualify for expanded food security coverage. That figure represents more than ration cards. It is the baseline from which parliamentary seats are apportioned, from which school construction and healthcare budgets flow, from which water and sanitation infrastructure is planned.

Dipa Sinha, a public health researcher and economist, who has tracked the census delay for years, is precise about the cascading damage. The problem, she argues, is not only the missing ration cards. It is that the census forms the basis for every sample survey that follows—health surveys, nutrition surveys, labour force surveys—all of which draw their sample frames from census geography.

Without a reliable census, the entire statistical architecture downstream is compromised. “By not having data for 16 years,” she says, “a period in which COVID happened and fertility rates changed dramatically across the country, we are effectively making policy blindly.” The National Food Security Act mandates coverage of 75 per cent of rural and 50 per cent of urban households—but the lists were locked to 2011. India’s population kept rising. The allocations did not.

Sinha is equally watchful about what Census 2027 includes and what it leaves out. The decision to enumerate jati data for the first time since 1931 she regards as long overdue—not merely as a concession to political demand, but as a necessary tool for seeing caste-based inequality in full. “It is not only about counting OBC populations,” she says. “It is about looking at how forward castes are doing, what the gaps are between different groups, and how sub-castes within OBC, SC, and ST categories are moving.” The methodological challenges, she notes, are surmountable—the state-level caste censuses in Bihar, Telangana, and Karnataka have already demonstrated that a long list of jatis can be enumerated in the field. The harder test will come after the data is released, when political claim-making begins and governments must decide whether findings actually translate into changed policy. So far, looking at those state exercises, that translation has not happened. “We will have to see how this goes nationally,” she says.

On the question of transparency, Sinha is careful. She does not believe individual-level microdata should be released—even anonymised village-level records, she points out, allow communities to identify who is who, and the Census Act’s privacy obligations exist for good reason. What she does insist on is that the aggregate tables the census has historically published—detailed breakdowns by geography, caste, and household type —must be maintained and released in full. “What the census has released up to 2011 is sufficient for research and policymaking,” she says. “I hope they maintain those same principles. There is really nothing much more that we need—we need what we have historically had, delivered on time and in full.”

On proposals to shift to a digital, continuous register, she is sceptical: not opposed in principle, but firm that India does not yet have the privacy protections, the regulatory capacity, or the institutional readiness to manage such a system across a country of this scale and complexity.

Dey proposes a community audit model modelled on the Jan Sunwai public hearings that MKSS pioneered in Rajasthan in the 1990s. Post census gatherings in villages and urban wards would let residents cross-check aggregate figures against lived experience: do the numbers match who actually lives here? What policies will follow? The idea borrows, in spirit, from the accountability framework that Telangana’s expert committee recommended for its own SECC—release everything, let civil society analyse it, let communities organise around what they find. Whether the national government will extend the same openness remains to be seen.

“No data means no accountability. It is vital that data is collected effectively, disseminated, and used so people can hold governments to account,” says Bharadwaj.

Whether Census 2027 delivers on its promise—technically flawless digital enumeration, caste data published in full, results within the same year as collection—will determine whether India closes the accountability gap that 16 years of absence opened. Whether it seizes the deeper ambition that Telangana modelled—a census not just of who lives here, but of how they live and why some live so much worse than others—will determine whether it is remembered as a bureaucratic exercise or a democratic one.

India has counted its people since the age of empire. A single state, lacking the luxury of waiting, showed what counting with purpose looks like. The nation now has the tools, the mandate, and, for the first time in sixteen years, the obligation. The question is whether it has the will.

Niharika Awasthi is a development communications professional with over a decade of experience in the sector. A keen social observer, she channels her love for reading and writing into impactful storytelling and advocacy.