The Mother of All Spreadsheets: What It Takes to Count a Billion-and-a-half People
The Pulse
The Mother of All Spreadsheets: What It Takes to Count a Billion-and-a-half People
There are programmes, and then there are programmes that define a generation of governance. India's Census 2027, underway since April 2026, is clearly the latter. With a budget of around $1.24 billion and three million officials working year-round to count the country's population, this is a major programme in its own right. No other entity in the world is currently running anything comparable in scale, complexity, or consequence.
Why this is a major programme
What makes India's census exercise worth examining through a programme management lens is not just its size. It is a confluence of ambitions and constraints, of political necessity and logistics at scale, of modern digital aspiration and the social fabric of a vast country. This is not merely a counting exercise. It is a strategic programme in the fullest sense: multi-phase, multi-stakeholder, politically charged, and carrying consequences that will shape the country's governance architecture for a decade.
The census meets every criterion of a major programme: the timelines, the cost, the complex stakeholder universe, and most importantly, the consequences that will outlast the programme itself.
India is meant to count its population every ten years, but this will be the first census in sixteen years, delayed first by the Covid-19 pandemic and then by a series of administrative issues. The long gap brings its own challenges. Every policy decision taken since, including welfare allocations, infrastructure investment, and poverty targeting, has been based on data that was already a decade out of date when the pandemic struck.
Then there is the digital transformation dimension. This is the first fully digital census, with every enumerator equipped with a government-issued mobile app capable of real-time data synchronisation. India joins Vietnam and Chile as one of the very few countries to attempt a fully paperless national census. The ambition is significant, but it also carries real challenges.
The strategic challenges
At the strategic level, the census leadership faces three challenges.
The caste challenge. The inclusion of caste enumeration, the first since 1931, is genuinely historic and deeply contested. The ruling Bharatiya Janata Party had long opposed caste enumeration, arguing it would deepen social divisions. It eventually conceded under sustained pressure from campaigners and caste groups. Caste is a political minefield, and the census leadership must navigate it while maintaining the neutrality that gives the data its authority.
The data-with-consequences challenge. Given the pace at which governments work, census data is unlikely to be published before 2029 although sources say 2027, which coincides with the world's largest election. The data collection exercise must be ringfenced from its political uses.
The scope challenge. Adding caste, expanding digital infrastructure, introducing self-enumeration, and linking to the National Population Register all reflect the ambition of the programme's sponsors. These additions are justifiable, but they also stretch the delivery model and create new potential failure points.
The operational challenges
The workforce challenge. Around 3.1 million enumerators and supervisors will be deployed, mainly schoolteachers and government officials. The training structure is staggering: 100 national trainers trained 2,000 master trainers, who trained 45,000 field trainers, who will train 3.1 million enumerators. At this scale, quality degradation is a risk at every level. Maintaining motivation, consistency, and accountability across a multimillion-strong workforce is an operational challenge no app can fully solve.
The digital-first challenge. Running a fully digital census across a country with deep connectivity gaps is a high-risk design decision. The self-enumeration portal was used by over 1.2 million households in the first 24 hours, a strong start. But the real test will come in rural areas, among the elderly, and in regions with limited connectivity. Beyond connectivity and adoption issues lie the risks of exclusion errors, data integrity failures, and authentication obstacles.
The geographic challenge. The sheer size of the country adds another layer. India has handled this well during major events such as general elections, but the challenge cannot be ignored. Enumerators must travel to around 640,000 villages and 7,000 towns, visiting every dwelling, whether in remote hills, tribal areas, or urban slums. Green or black swan events could delay the effort considerably. Even some urban clusters may prove difficult to reach.
The tactical challenges
At the tactical level, in the day-to-day collection, the challenges become even more granular.
Household complexity. Urban India poses real definitional problems. Informal settlements, multi-generational joint households, migrant worker accommodation, and gig economy workers who move between cities all complicate the simple act of counting. Who counts as a resident? Where does a migrant worker belong? What about undivided families, or long-term visiting families with a permanent base elsewhere? These are questions enumerators must resolve consistently, at speed, across millions of visits.
Caste classification at scale. Creating a standardised national caste code directory is complex, given the regional, linguistic, and sub-group variations in caste identities. The census will list individual jatis based on self-identification rather than mapping to central or state lists. In theory, this avoids one set of disputes. In practice, it creates a classification problem at the front end, where thousands of self-reported caste names will need to be reconciled into a usable taxonomy.
Community trust and participation. Not everyone wants to be counted, and not everyone will answer honestly. Marginalised communities, including tribal populations, undocumented migrants, and minority groups, may have historical reasons to be wary of enumeration. The programme's public communication strategy, described as offering "last-mile engagement", will be tested hardest precisely where the stakes for accurate data are highest.
How it compares to other countries
India | UK | Vietnam | |
|---|---|---|---|
Year of census | 2027 | 2021 | 2019 |
Coverage | ~1.4 billion | ~60 million | ~96 million |
Response rate | TBD | 97% | 99.95% |
Online response % | TBD | 88.9% | First attempt |
No. of enumerators | 3.1 million | ~30,000 | ~147,000 |
Turnaround time | Target: 2027 | ~15 months | ~3 months |
What this programme team can do well
Several things remain within the programme leadership's control.
Keep data and politics separate. Establish a firewall between data collection and data use. Publish a clear data governance framework, make the methodology transparent, and engage independent academic institutions in quality assurance. The moment the census is seen as politically manipulated, its utility is diluted.
Invest disproportionately in hard-to-reach populations. Standard enumeration will likely undercount communities that are already data-invisible: nomadic groups, slum dwellers, undocumented migrants, tribal populations. They need specialist approaches beyond the standard training cascade. The cost of undercounting them is not evenly distributed; it concentrates in the policies meant to serve them.
Build quality assurance into the workflow. Real-time data from the mobile app allows the team to spot geographic gaps, inconsistencies, and anomalies during collection rather than months later. That data should be used actively during enumeration, not treated as a processing-stage problem.
Design the caste taxonomy before collection, not after. Collecting self-reported jati makes methodological sense, but creates a classification challenge that will take years to resolve if not designed in advance. A transparent caste coding framework, developed with sociologists and community representatives before Phase II, will save processing time and reduce scope for politically motivated reclassification.
Treat this as the foundation for the next census. The digital infrastructure being built for Census 2027 is a beginning, not an end. The team should document the architecture, failure modes, and workarounds as they go, so that Census 2037 starts from a far stronger base.
What this tells us about major programmes
India's census is, in a way, a portrait of what all major programmes eventually become: a technical exercise overshadowed by its own political and social consequences. The challenge for programme leadership is to hold operational delivery together while the strategic environment shifts around it.
The census is not just a measurement exercise. It is a political one. That is true of most major programmes at sufficient scale. The skill lies in doing the technical work well enough that the political questions can eventually be answered with evidence rather than assertion.
The world is watching the largest administrative exercise in human history unfold in real time. Whether it succeeds will depend less on the technology than on the human decisions made at every level of the system, from the Cabinet committee that sets the scope, to the schoolteacher who knocks on a door in a village no map accurately shows.
