Every day for about a month, M. Soumina, a government-school teacher, began her mornings by knocking on the doors of apartments in a multi-storey residential complex in East Delhi’s Mayur Vihar.

Although the school where she taught was closed for the summer break, Soumina had her task cut out for her after being appointed an enumerator for the ongoing Census. The exercise relies on about 33 lakh government teachers and other enumerators, who will spend nearly a year surveying an estimated 1.4 billion Indians on housing, amenities, and demographics.

For Soumina, the work meant spending hours going door to door in a brutal Delhi summer, with temperatures consistently hovering between 40°C and 42°C. She surveyed 150 flats spread across 17 apartment blocks and administered a 33-question survey to each household within 26 days.

“I was making those visits daily from May 19,” said Soumina. “I had an app on my phone to record the data. Before asking the questions, the households also have to be mapped and geotagged on the app.”

Other than her smartphone, a black bag, a QR-enabled identity card, and a white cap with the Census logo to protect her from the heat, Soumina’s companion during these visits was her husband, who works in the production unit of a newspaper. She said she felt safer with him around.

Like Soumina, enumerators working on the Census report facing several challenges in the field, including heat, connectivity issues, and safety concerns. Though the exercise is being conducted through a digital system designed for real-time monitoring, field workers said conditions on the ground and supervisory instructions are affecting data collection. Many of them have also flagged these issues on social media.

A massive exercise

The questions that enumerators asked are part of the Houselisting and Housing Operations (HLO), the first phase of the Population Census 2027. The questionnaire was notified by the Registrar General and Census Commissioner of India on January 22.

The 2027 Census marks several firsts: it is India’s first fully digital Census, the first to collect caste data, and the first to allow residents to self-enumerate. The exercise was originally scheduled for 2021, but was postponed because of the COVID-19 pandemic. As a result, India is conducting its population count six years later than the decennial schedule, with governments continuing to rely on 2011 Census data for policy planning and welfare programmes.

The HLO phase involves the collection of data on housing conditions, household composition, basic amenities, and assets to assess living standards. It captures information on the structure of dwellings; access to water, sanitation, and energy, ownership of consumer goods, and the number of people living in a household. It also collects residents’ mobile numbers for official communication.

The HLO phase will lay the foundation for the second phase of the Census — Population Enumeration — scheduled for February 2027. In Ladakh and the snow-bound regions of Uttarakhand, Jammu and Kashmir, and Himachal Pradesh, both phases have been set a deadline of September 30, 2026, due to weather constraints.

The second phase will enumerate caste and collect other detailed individual-level data, including religion. Each enumerator gets an honorarium of ₹25,000 for completing the task. So far, 23 States and Union Territories have completed the first phase.

Meeting strangers

With several occupants being working professionals, Soumina said it was difficult to get them to answer questions during the day.

“Most residents leave for work around 8.30 in the morning. So I visited homes between 7 and 7.30 with my husband. While some were accommodative, others asked us to come again later or skip the exercise altogether,” she said.

If a house was locked, she would go back again and again until she got the data, Soumina added. “Sometimes the visit extended into the evening, so I would again go with my husband.”

Like her, many female enumerators, mostly government school teachers, said they asked their husbands or other male members in their families to accompany them during field visits.

Meenu Verma, a primary school teacher from Rae Bareli, Uttar Pradesh, said that her husband and brother-in-law helped her during the exercise. “I was assigned 172 houses,” she said. “My husband and I used to wake up at 4 in the morning and reach the assigned village by 7. From 7 to about 11, we would find people in their homes, but after noon, due to the heat, the entire village would be deserted.”

She said they would then hesitate to knock on doors. “Many of them had no awareness about the Census, so we had to first explain it them. I want to request the government to appoint enumerators in pairs. Or a male colleague should accompany a female enumerator. Not everyone is polite. And in villages, houses are spread afar,” Verma said.

On May 21, a First Information Report (FIR) was registered against a Delhi resident for allegedly harassing a female Census enumerator while she was on duty. On June 13, a Delhi court convicted the accused, Sandeep Panwar, 43, for the crime. Earlier, on April 25, a female Census enumerator was allegedly denied entry into an apartment building in Bengaluru where an unidentified individual allegedly used abusive language against her, following which an FIR was registered against him.

Dinesh Kumar Dwivedi, a Hindi lecturer at a government school, who was Soumina’s supervisor, said six enumerators were reporting to him. “Fetching data from slum clusters was easier than seeking information from multi-storey apartments,” he said.

In Haryana and Delhi, Census enumerators, mostly government school teachers, complained that they faced hostility and were denied entry by residents of high-rise buildings and residents’ welfare associations (RWAs). The district administration of Gurugram wrote letters to RWAs saying that according to Section 8(2) of the Census Act, 1948, every resident is “legally bound” to answer Census-related questions to the best of their knowledge or belief.

“The slum residents were hesitant initially,” said Dwivedi, who collected data as an enumerator during the 2011 Census. “They thought this was an exercise to demolish the slums. When they understood its purpose, they were more than forthcoming. Some thought that participating in it would make them eligible for government schemes. Many others came to us with documents, asking for help in applying for welfare schemes.”

Dwivedi recalled that in 2011, the process was tedious: all the data had to be noted down on paper, which could get lost or damaged. He said residents were reluctant to share information on assets as they feared tax raids.

The digital divide

Beyond personal safety, the digital divide in India has also affected fieldwork. In metros and towns with better mobile connectivity, enumerators entered data directly on their phones. Elsewhere, they said they recorded details on paper or in notebooks and uploaded them on the app after going back home.

Read: Census 2027: heat, locked doors pose challenges for survey deadline

According to the Comprehensive Modular Survey: Telecom, conducted from January to March 2025 as a part of the 80th round of the National Sample Survey, 83.3 % of rural households and 91.6% of urban households had access to Internet. The total number of households surveyed was 34,950 (19,071 in rural areas and 15,879 in urban areas) and the total number of people was 1.42 lakh (82,573 in rural areas and 59,492 in urban areas). In the survey, a person was considered to have “used the internet” if they had accessed it at least once in the past three months.

“Connectivity is a huge issue, especially in rural areas. Once an enumerator has logged in, it is difficult to make changes, and if the network is poor, the data can get corrupted,” said an enumerator from U.P.

Enumerators use their own phones to collect the data and have been asked not to delete the apps during the enumeration phase. However, they said the app crashes often.

On the Google Play Store, the Census 2027 Houselist App, which has more than 10 lakh downloads, has a rating of 2.9. It can only be downloaded by authorised users — enumerators, supervisors, and charge officers. Of the 4,570 reviews submitted, 2,054 users gave it a one-star rating, while 1,735 rated it five stars.

Facing FIRs

On the outskirts of Delhi, in U.P.’s Noida, Sector-6, Indu Prakash Singh, charge officer, who oversees the work of enumerators and supervisors, set up a camp office at the Indira Gandhi Cultural Centre to monitor Census-related work. Officials with technical expertise were stationed in a large hall, with computer screens placed on desks arranged in a semicircle.

Singh’s job was to ensure the smooth conduct of the exercise, which entailed surveying 2,309 house-listing blocks (HLBs), each comprising 150-180 households with a population of 800-1,000. The exercise ran from May 22 to June 20 in U.P.

Though Census activities are usually carried out by government employees, in U.P., private-school teachers were also enrolled to tide over the shortage of government school teachers.

“The district administration in Noida issued directions to engage teachers from private schools for Census work,” Singh said.

“But some of them moved the Allahabad High Court, which ruled in their favour, and 50% of the 1,000 teachers who were trained backed out at the last moment. We were forced to engage NGOs and other private people for the task,” said a government official in Noida.

In Maharashtra, over 500 private unaided and minority schools moved the Bombay High Court in May against notices, appointment orders, and coercive steps such as FIRs issued against their teaching and non-teaching staff for failing to report for Census duties.

On May 22, the Court granted interim protection to the school associations and observed that Section 4A of the Census Act applies only to “local authorities” and prima facie the Census Act and the Census Rules, 1990 do not impose any statutory obligation on private unaided schools to provide their staff as enumerators or supervisors.

There have been several instances of enumerators facing FIRs for not taking up the assigned work. Though the total figures are difficult to assess, at least two FIRs were filed in Haryana against 10 enumerators, while a case each was filed in Noida and Greater Noida against more than 80 enumerators on June 3.

The National Crime Records Bureau reports for 2010 and 2011, when the first and second phases of the Census were last conducted, do not mention any cases registered under the Census Act.

Enumerators took to social media to flag issues such as heat and connectivity problems. They also reported witnessing stark income inequality, and described pressure from superiors to edit data in the name of correcting discrepancies.

Editing data

Earlier, it often took months or even years to process Census forms. Mobile apps have now enabled real-time data monitoring through the Census Management and Monitoring System (CMMS) portal. Access to the data entered via the app is available not only to enumerators, but also to supervisors and charge officers. Enumerators said they were asked to change data, particularly for questions related to sanitation, electricity, and LPG connection.

A school teacher in Delhi said during training, for a question relating to “access to latrine”, they were asked not to select the “no” option. This is used as a category for households without access to a latrine, and indicates open defecation. The “Yes” option is further broken down into “Exclusively for household use only (1),” “Shared with other households (2),” and “Public latrine (3).”

“During training, we were told to fill in the data observed by us,” said an enumerator from Rajasthan. “On the field, many residents in my enumeration block said they defecate in the open. When I submitted the data, I was asked by my supervisor to revisit the homes and revise the data to show that the households had access to a latrine, even if a public toilet was anywhere in the vicinity.”

Another enumerator from U.P. said they were also asked to edit data on the main source of drinking water. The options include tap water from treated source/untreated source, well, hand pump, tubewell/borehole, spring, river/canal, tank/pond/lake, packaged/bottled water and other sources. Under the Jal Jeevan Mission, one of the flagship schemes of the Bharatiya Janata Party government, nearly 81.97% households in India have tap water connection.

“When I filled untreated tap water as an option, the supervisor and the charge officers asked me to change the entry to tap water from treated source. What is the point of door-to-door enumeration then,” asked the enumerator.

The question on the availability of a kitchen and LPG/PNG connection were also edited, the enumerator added. “Even if the house has an LPG connection in their name, they were using cow dung cakes or wooden logs to cook as they had not refilled the cylinders for several months. If we mentioned firewood or charcoal as the response, which is what we observed, we were asked to change the entry to show that the main fuel is LPG,” said an enumerator from Rajasthan. According to the Press Information Bureau, national LPG coverage rose from 55.9% in April 2014 to 107.2% in April 2026. Consumers grew from 14.51 crore to 33.39 crore in the same period.

On June 2, the Director of Census Operations, Rajasthan, wrote to all district functionaries saying “during the analysis of field data collected so far, some discrepancies have been noticed”. Charge officers were instructed to “verify the block-level data through the CMMS portal, in accordance with the actual field situation.”

Around 11 a.m. on June 12, Manish Kumar, a government employee, went to a building in Noida’s Sector-2. An open drain flowed on the side. Several doors were locked since many of the occupants, largely migrants, were away at work. In the courtyard of the three-storey building, a woman was washing bedsheets in a corner using two tubs of water. A man was taking a bath in a narrow veranda outside his one-room accommodation, which he shared with six other members of the family.

Kumar knocked on each door to ensure that every household was enumerated. The rooms that were open mostly had women and children inside. While some residents responded immediately, others were suspicious of Kumar. He noted down the details, stepped out of the building, and then logged the information into the Census app on his phone.

He said he would return in the evening and over the weekend to reach residents whose doors were locked. “A household must be labelled vacant or locked only after multiple visits,” he said.