Counting India: The first digital census and its blindspots

India’s Census 2027, the country’s sixteenth decennial census, marks a decisive break from the past. For the first time in the nation’s census history, the exercise is set to be fully digital. Gone are the dog-eared record books and the handwritten household registers that enumerators have carried door to door since 1881. In their place are a GPS-enabled mobile application, satellite-based geotagging of buildings, and a citizen-facing self-enumeration portal at se.census.gov.in. Counting over 1.4 billion people across two phases, this census aims to produce the most detailed demographic dataset India has ever generated.

By incorporating new questions on digital connectivity, gender identity, and climate-induced migration, the census moves well beyond the conventional demographic headcount. The data collected will serve as the foundational evidence base for welfare targeting, infrastructure planning, parliamentary delimitation, resource allocation, and disaster management. It will inform virtually every major policy decision over the next decade.

The census is to unfold in two phases. Phase I, the house listing and housing census, runs from April 1 to September 30, with each state and union territory conducting its 30-day listing window within this broader period. Phase II, the Population Enumeration, is planned for February 2027.

Phase I carries the more significant structural innovation. The 30-day window is divided into two distinct fortnights. The first 15 days are reserved for self-enumeration, during which residents may fill in household details using a laptop or mobile phone and submit them via a registered mobile number through se.census.gov.in, which supports 16 languages. The remaining 15 days are for enumerators to visit households, verify submitted data, and record information for those who have not self-enumerated. Upon submission, a unique Self-Enumeration ID is generated, which the household shares with the enumerator during the field visit.

This bifurcated approach reduces the burden on field enumerators, improves data accuracy by placing primary responsibility on the household itself, and signals a broader shift in how the State relates to citizens in administrative exercises.

It all sounds straightforward enough. Yet the process is not quite as simple as it appears on paper. The self-enumeration portal asks residents to pin their household location on a digital map, and enumerators subsequently geotag buildings using the Janganana mobile application. For a significant portion of the population, confidently identifying one’s own house on a digital map is not easy — a limitation that gets more pronounced in densely built urban areas, where satellite-based positioning often struggles with signal obstruction from tall buildings, narrow lanes, and closely packed structures. The assumption that location technology will perform reliably and uniformly across India’s varied urban fabric deserves far more scrutiny than it has received in official communications.

The house listing questionnaire contains around 33 questions. It covers building materials such as walls, floors, and roof types, along with ownership status, drinking water sources, cooking fuel, household size, and the social category of the head of the household. It also captures asset ownership across radio, television, telephone, computer, laptop, Internet access, and two-wheelers.

This is where one must pause. The asset matrix, while comprehensive by the standards of an earlier era, appears inadequate for 2026. The absence of questions on smartphone ownership as distinct from basic telephone access, on digital payment usage, or on electric vehicles reflects a questionnaire that has not fully kept pace with India’s rapidly evolving socioeconomic reality. For a dataset that will inform policy well into the mid-2030s, such gaps risk producing analyses that are outdated even before the final report is published.

Despite these reservations, the case for self-enumeration is strong. The process takes approximately 20 minutes. More importantly, it places the accuracy of your household’s data in your own hands. In previous censuses, enumerators occasionally made assumptions about household composition, social category, or asset ownership, sometimes due to time pressure and sometimes due to language barriers. Each such error, multiplied across millions of households, distorts the policy conclusions that flow from the census.

There is, however, one operational constraint worth flagging. Once the household head registers a mobile number on the portal, it becomes linked to the geocoded household record. Residents should be aware of this privacy consideration before they register, as the official communication on this aspect has so far been limited.

The self-enumeration window varies by state, running in 15-day slots between April and September. Early schedules have reportedly included Karnataka, Goa, Odisha, Sikkim, Mizoram, Lakshadweep, Andaman & Nicobar, and parts of Delhi in the first half of April. Other states are expected to follow in subsequent windows, though final schedules may vary.

All it takes is 20 minutes. Census data shapes where the government builds schools and hospitals, how it allocates funds, and how constituency boundaries are drawn for the next decade. Citizens can leave that record to someone else’s judgment, or can fill it in themselves and get it right. The choice, for once, is entirely citizens’.

(The writer is professor and head of the Geospatial Research Programme,
Takshashila Institution, Bengaluru)

(Disclaimer: The views expressed above are the author's own. They do not necessarily reflect the views of DH.)