VISTAR must pair better data with safeguards and state capacity in India

More steps needed to improve official data collection

The Union Cabinet’s recent approval of the ₹4,371 crore Viksit Bharat Initiative for Statistical Transformation, Analytics and Research (VISTAR) recognises that India’s statistical infrastructure must keep pace with the changing economy. It needs information that is timely, reliable, comparable, and granular enough to capture differences across sectors and regions. VISTAR, which will run from 2026-27 to 2030-31, offers an opportunity to address longstanding weaknesses. The case for more frequent data collection is particularly strong at subnational level. A study published by the Reserve Bank of India (RBI) in its October 2026 bulletin illustrated the problem. Official annual gross state value-added estimates typically become available with a significant lag, limiting the ability to track regional economic conditions. The study developed a framework for estimating quarterly growth in Gujarat, Maharashtra and Karnataka by combining satellite-based night-time lights with Google Trends data. Changes in luminosity offer a proxy for economic activity, while online searches relating to jobs, consumption, credit and housing provide behavioural signals. Such approaches cannot replace official estimates, but show how alternative sources can bridge information gaps.

 

VISTAR’s proposed expansion from 43 to 59 surveys during the 16th Finance-Commission cycle is significant. So are the planned headline index of services production and the index of unincorporated sector enterprises. The eighth economic Census, covering non-farm establishments and workers down to gram panchayat and ward levels, could provide a more comprehensive picture of the enterprise landscape. Its proposed National Statistical Business Register could offer a sampling frame for future surveys. The proposed use of mobile-phone records, scanner data, and ecommerce information is another promising feature. These sources can provide frequent signals and reduce the burden on respondents. Artificial intelligence and machine learning could assist data processing, validation and analysis, while common standards and sharing protocols could reduce duplication across departments.

 

However, technological sophistication should not be mistaken for statistical quality. Digital transactions and online searches do not represent all households or businesses; night-time lights, too, are an imperfect proxy for production. Estimates derived from such sources must be tested against representative surveys, while making methods, limitations and revisions transparent. Privacy safeguards must also govern the use of personal and commercially generated data. Equally important is the support for strengthening the statistical component. National estimates depend on the quality of information collected and compiled by states. Support for district-level and million-plus city estimates, state- and district-level gross domestic product, and state-level industrial production could reveal disparities concealed by state or national averages. Efforts must be made to build lasting capacity at different levels. VISTAR should also create opportunities for researchers to examine data, test methods, and develop indicators. Wider access to machine-readable datasets, subject to privacy protections, and timely publication of survey results would encourage independent scrutiny and strengthen public confidence. Better data collection and dissemination will help both policymakers and businesses to make informed decisions and improve outcomes.