Administrative records can reshape policy
Every few years, India’s official statistical machinery undertakes one of the largest data collection exercises in the world. Surveyors travel across the country to record household incomes, employment , health outcomes, sanitation practices and educational attainment. These surveys, from the National Family Health Survey (NFHS) to the Periodic Labour Force Survey (PLFS) have long been the backbone of evidence-based policymaking. Yet, by the time the data are cleaned, analysed and published, the country they describe has often moved on. Today, evidence about India is also being generated continuously as a by-product of routine digital transactions.
Every UPI transaction, GST return, welfare payment, school enrolment and hospital registration leaves behind a digital administrative record. Individually, these records exist to deliver public services. Collectively, they provide governments with a continuously updated picture of the economy and society. This points to an important shift within official statistics: the growing use of administrative data as a complement to surveys and censuses. The Ministry of Statistics and Programme Implementation has described administrative data as data generated as a by-product of routine administrative processes, offering granular insights that help move governments towards more continuous and comprehensive data systems.
Administrative data should not replace surveys; it should complement them. Surveys remain indispensable for measuring outcomes that routine government records cannot capture such as perceptions, aspirations, unpaid work and household well-being. Administrative data, however, offers something surveys often cannot: timeliness, scale and operational detail. Used together, they can move policymaking from retrospective assessment to real-time governance. Built on the Jan Dhan-Aadhaar-Mobile (JAM) architecture, platforms such as the Goods and Services Tax Network (GSTN) and the Unified Payments Interface (UPI) generate high-frequency indicators of........
