Counting citizens or capturing data?

Conducting a census is never easy in a country as vast and populous as India. Yet, over the decades, India has built a robust system to carry out this massive exercise, navigating its complexities with relative efficiency.

Until now, neither successive governments nor the general public have found the process to be complicated, confusing or burdensome. That may change with the upcoming 2027 census, which is already six years behind schedule. The little we know about it has only added to the uncertainty.

What’s even more worrying, some of the information that has emerged hints at potential new risks. Following the first Union cabinet meeting after the Pahalgam terrorist attack in April 2025, it was announced that the next census would also include a caste count.

However, no details were provided about how this would be implemented. A press release from the MHA (ministry of home affairs) on 4 June merely outlined part of the census timeline. Then, on 16 June, the Registrar General and Census Commissioner issued a notification in the official gazette, which again mentioned the dates but no other specifics. Interestingly, a day before this notification, Union home minister Amit Shah had chaired a meeting of senior officials to discuss the census.

The government’s press release included only procedural details—such as the number of personnel and supervisors who will be involved in the exercise. It also confirmed that this would be India’s first digital census, with data collection taking place through a mobile application.

Citizens would also be allowed to submit their details through a self-enumeration option. However, the press release remained silent on the potential risks and challenges of conducting a digital census in a country with significantly poor digital literacy, digital divides and data privacy concerns.

The concerns surrounding the upcoming census are serious—but before going into these risks, it is equally important to examine what has not been said. Significantly, after the 30 April press briefing, the government has remained completely silent on the issue of caste census prompting Congress leader Jairam Ramesh to ask whether the government had had a change of heart.

Praveen Chakravarty, head of the All India Professionals’ Congress, who was involved in the caste survey process in Telangana, told National Herald, “We have no idea what the government plans to do. Nothing official has been shared so far.”

The contours of the proposed digital census are becoming a fresh cause for concern. India’s past experiences with digital data collection have not been particularly reassuring, and the government has made no real effort to address growing fears around data privacy and protection. The rollout of India’s first fully digital census raises new and serious questions about intent, capability and speed.

So, what exactly does a digital census entail? Cyber law expert Pawan Duggal explains that unlike earlier censuses conducted with pen and paper, this one will use handheld devices—mobiles or tablets. Enumerators will record responses to questions like: How many people live in your home? Is the house made of permanent (pucca) or temporary (kutcha) material? Do you have an LPG connection?

In addition, GPS-tagging and geo-fencing of respondents will also be done. This marks a fundamental shift in the nature of the census. It no longer stops at counting people or households. It becomes a tool for collecting extensive personal data, without explicit consent.

In effect, the digital census could gather and store not just the information people choose to give, but also several layers of sensitive personal details—raising grave concerns over privacy, surveillance and State overreach. A marked change in how data is collected, and how it is used. In today’s digital economy, the most valuable—and vulnerable—data includes not only personal information but also GPS locations.

India’s track record with digital data collection has not been reassuring. The Aadhaar system has faced repeated data breach allegations. Apps like Aarogya Setu and DigiLocker have drawn flak for poor user experience and inconsistent data protection.

Duggal warns that safeguarding data will be a major challenge, not only because of commercial threats but also potential political misuse. “When we don’t even have a comprehensive cybersecurity law,” he asks, “how can we expect the digital census to prioritise data protection?”

Professor Jagdeep Singh Chhokar of the Association for Democratic Reforms echoes this concern. “They can’t even manage Aadhaar,” he says. “Aadhaar data is already being bought and sold in the open market. How can we trust them to protect digital census data?” He raises a deeper concern: the integrity and intent of the government. “This isn’t just about data breaches,” he argues. “It’s about whether we can trust the government not to misuse this information.”

Economist Jayati Ghosh recently expressed a similar fear. “At times,” she said, “it feels like the Modi government fears data—except when it’s in their own hands.” This is the crux of the problem: the danger that data gathered in the digital census exercise could be used to microtarget voters, build digital profiles or shape election narratives.

Reports from previous elections have already shown that detailed data from welfare schemes found its way into the hands of the ruling party. According to the 16 June press release, over 31 lakh enumerators and supervisors will be deployed for doorto- door data collection. Some reports indicate that around 4,600 personnel will be assigned solely for training purposes.

However, Chhokar points out a significant flaw in this setup. Simply knowing how to operate a computer, tablet or smartphone does not ensure proficiency in data entry. “Data entry is a specialised task,” he says. “In the past, trained professionals handled the process. If any discrepancies arose, the original paper records could be crosschecked. But in the digital model, there is no fallback.”

While proponents argue that going digital will reduce human error, the opposite may occur if data collection is rushed or handled by inadequately trained personnel. Fear of manipulation of the data at different levels will remain a constant concern. Duggal says this concern is compounded by the digital divide which could result in underreporting, especially among vulnerable and marginalised communities.

One of the few touted benefits of the digital census is speed. But that too seems a dubious promise. The 4 June government press release stated that groundwork for the 2021 census had already been completed before the Covid-19 outbreak. If so, the delay in launching it is difficult to justify.

A stark contrast can be seen in Telangana’s recent caste survey. Chief minister Revanth Reddy took office on 7 December 2023. Less than two months later, on 4 February 2024, his cabinet approved the caste survey. In the months that followed, the state completed key steps: passing a resolution in the assembly, consulting experts and stakeholders, preparing the survey blueprint and designing the questionnaire.

Notably, while the 2011 national census had just 30 questions, Telangana’s survey featured 75. Yet, it was completed well before Reddy’s government marked its first anniversary. If Telangana could plan and execute a complex, expansive survey in less than a year, why is the national digital census still stuck at the starting line?

Chakravarty puts it plainly: “With the kind of technology available now, there is no justification for the delay we are witnessing. The process should be far quicker.” In an era of advanced information technology, conducting a census is no longer as daunting or time-consuming, and yet the government seems to be moving at a pre-digital pace despite the technological tools at its disposal.