Discuss the governance and privacy concerns of collecting caste and Aadhaar-linked information in the Census. Suggest safeguards.

Census 2027 is India's first digital Census. In its 40-question Population Enumeration, around 34 lakh enumerators will record caste and Aadhaar details on a mobile app running on their own phones [1]. The Cabinet has approved ₹11,718.24 crore for it [2]. Because of this, the main test is now how well the data is governed, not only how well people are counted.

Governance concerns

• Training cascade: instructions pass from the RGI to Directors of Census Operations, then to field trainers, then to enumerators. Careful wording for sensitive questions can get simplified at each step [1].
• Diluted accountability: since 16 October 2026, private school teachers and members of societies and trusts can also be enumerators. This weakens the government chain of command [1].
• Classification risk: there are thousands of sub-castes and synonyms. Without neutral questioning, answers may be misclassified or respondents pressured. This affects reservation and delimitation debates [1].
• Time pressure: each household needs 20–25 minutes, which tempts enumerators to take shortcuts [1].

Privacy concerns

• Personal devices: caste and Aadhaar answers are stored, at least for a while, on phones the State does not control. UN guidance expects the census budget to pay for official devices [6].
• Profiling risk: Aadhaar could serve as a common key linking caste data to other databases. This would fail the necessity-proportionality test of Puttaswamy (2017) [4].
• Gaps in the law: the Census Act, 1948 keeps census records confidential [3], but it was written before digital data. The DPDP Act, 2023 gives the State wide exemptions [5].
• Trust deficit: if people fear misuse, they may refuse to answer or give false answers.

Safeguards

• Legal: notify rules that ban linking census records with Aadhaar-seeded databases, and set rules on purpose limitation and data retention.
• Technical: lock and encrypt the app, delete data from the phone automatically after upload, and mask Aadhaar numbers.
• Procedural: give late-joining enumerators a short catch-up module on the caste and Aadhaar questions and on PRAMAN's Pehchaan (identity) norm [1]. Supervisors should check data for patterns while fieldwork is on [6].
• Citizen choice: expand self-enumeration, so households enter sensitive details themselves [7].
• Accountability: make all non-government enumerators take a secrecy oath, and have an independent body audit data handling.
• Classification: an expert panel should publish standard caste lists for each State, so answers are coded the same way everywhere.

Caste and Aadhaar data can sharpen welfare targeting. The RGI calls the Census the "empirical foundation" of Viksit Bharat [1], but that only works if citizens trust the process. Building privacy into the law, the app and the training would honour the privacy guarantee under Article 21. It would also make Census 2027 a model of data governance that respects rights.