Before a state can support a farmer, it has to know that the farmer exists — who they are, what land they cultivate, what they grow, and where money should be sent. For much of India's history, that knowledge lived in paper land records, local registers and the discretion of intermediaries. The result was a system in which benefits leaked, eligibility was contested and many farmers simply did not appear in the data on which policy depended.
From payments to profiles
The first layer of change came through what is often called the JAM trinity: Jan Dhan bank accounts, Aadhaar identity and mobile phones. Together they made direct benefit transfer possible at scale. PM-KISAN's 23rd instalment, released in June 2026, reached 9.44 crore farmers without the money passing through intermediaries — an administrative undertaking that would have been hard to imagine two decades ago.
The second layer is more ambitious. The Digital Agriculture Mission, including AgriStack, was approved in September 2024 to build registries that link a farmer's identity to land records and to seasonal data on the crops sown, gathered through a digital crop survey. In principle, this lets a farmer be verified once rather than proving themselves afresh for every scheme, and allows insurance, credit and advisory services to draw on the same authenticated record.
The risk of being seen incorrectly
Being visible to the system is powerful. Being invisible, or visible incorrectly, can be costly in new ways. When eligibility is determined by a database, the errors in that database become the terms of inclusion. Several groups face particular exposure:
- Tenants and sharecroppers, who cultivate land they do not own, are often absent from land records, and work under arrangements that are frequently informal.
- Women farmers, who carry out a large share of farm work but are less likely to hold land titles in their own names.
- Households with record errors — mismatched names, outdated mutations, unresolved inheritance — for whom a spelling discrepancy between two databases can halt a payment.
When eligibility is determined by a database, the errors in that database become the terms of inclusion.
A digital system can also shift the burden of proof. Under paper administration, a local official could exercise judgement. Under automated verification, a farmer rejected by the system may not know why, whom to approach, or how to correct the record. The design questions that matter are therefore as much institutional as technical: how consent is obtained for the use of personal and land data; who may access that data and for what purposes; and whether grievance redress is fast, local and empowered to override the database when the database is wrong.
There is a further question of what digital identity is ultimately for. A registry built only to deliver subsidies more efficiently will make the existing welfare system leaner. A registry that farmers can use — to obtain credit on the strength of their crop history, to demonstrate their output to buyers, to receive advice suited to their plot — could change their position in markets. The two uses are compatible, but they are not the same, and the second depends on farmers having meaningful control over their own records.
Digital agriculture will ultimately be judged less by how many farmers it has registered than by how quickly it corrects the record of the one it got wrong.