For a knowledge-driven economy, India must fix how it measures its progress on R&D

The biennial R&D Statistics survey by the Department of Science and Technology (DST) through its National Science and Technology Management Information System (NSTMIS) is the sole official source of national data on R&D activity.

Since 1973, it has provided information on expenditure, personnel, patents and institutional performance across roughly 8,000 organizations. Yet, this once robust system is now struggling to capture the growing complexity and scale of India’s innovation landscape.

The latest NSTMIS round, launched in December 2024, highlights a chronic weakness—low participation by private firms. While government laboratories and research institutions achieved a response rate of about 73%, only about 35% of firms registered with the Department of Scientific and Industrial

Research and fewer than one in 10 multinational corporations took part. Industry representatives cite several reasons: ambiguity in defining R&D expenditure, lengthy reporting formats and concerns over confidentiality. The result is a data-set that fails to reflect the true extent of private-sector research.

Policymakers rely on these figures to design incentives and measure progress. Without credible data, it is impossible to judge whether fiscal measures such as weighted tax deductions or targeted funding are achieving their intended outcomes.

The Global Innovation Index (GII) 2025 ranks India 38th, up from 66th in 2013, making it the best performer among lower-middle-income economies. India now ranks first globally in ICT service exports and fourth in late-stage venture capital activity, reflecting a vibrant digital and entrepreneurial ecosystem. Yet, beneath this progress lies a sobering reality: total R&D expenditure has stagnated at around 0.64% of GDP for over two decades—among the lowest in the G20.

The NSTMIS framework has several structural weaknesses. Its definitions of R&D are not fully aligned with the OECD’s Frascati Manual, the global standard that distinguishes genuine research from routine development or market research. This conceptual ambiguity leads to inconsistent reporting and undermines international comparability.

The survey also omits basic quality indicators such as response rates and coverage or non-response biases, and fails to provide cross-classified data that can show funding flows between sectors. Such performer–funder metrics are essential to assess public–private collaborations and the real structure of R&D funding.

Time lags also limit usefulness. The most recent report released in 2023 covers spending only up to 2020-21. Without timely data, it is impossible to evaluate if initiatives such as the Anusandhan National Research Foundation (ANRF) or the production-linked incentives are having any measurable effect on R&D intensity.

Efforts to embed R&D into mainstream economic policy have also been uneven. In 2019, the Office of the Principal Scientific Adviser recommended integrating R&D expenditure into the Union budget and the Economic Survey. While the creation of ANRF is a welcome institutional development, the budget still does not present consolidated national R&D accounts.

At the state level, only a few governments—Kerala and Odisha—publish dedicated R&D budget statements. This fragmented approach keeps research expenditure largely invisible in fiscal planning and signals to private firms that R&D remains peripheral to India’s economic strategy.

Reforming the R&D data system should thus be a policy priority.

First, create a fully digital submission portal with AI-assisted validation tools and mobile-friendly interfaces. Automated error detection and pre-filled fields linked to corporate filings would reduce reporting burdens and improve accuracy.

Second, the NSTMIS must adopt greater transparency by publishing detailed methodological notes to build credibility and align with international norms.

Third, R&D data should be linked to corporate filings maintained by the ministry of corporate affairs under its MCA21 system, ensuring consistency between fiscal and innovation statistics.

In the long term, India should consider a statutory framework for R&D reporting, modelled on the Annual Survey of Industries under the Collection of Statistics Act of 2008. The mandatory but incentivized reporting systems of OECD countries have played a role in the large share their private sector firms have in total R&D expenditure.

Our innovation ambitions, from developing indigenous semiconductors to advancing clean energy and artificial intelligence, depend on sound and timely data. Without reliable R&D statistics, policy decisions risk being driven by assumptions. Upgrading the R&D survey system would improve data quality and strengthen trust between the government and industry.

This is not a mere statistical exercise; it is a foundational step towards making India’s innovation policy evidence-based, credible and more supportive of a knowledge-driven economy.

These are the author’s personal views.

The author is a visiting professor, Centre for Development Studies, Trivandrum, and Ahmedabad University.