Why India's CSR data is rich and its intelligence is poor

The MCA dashboard records the spending of 18,000 companies. Every rupee of ₹34,909 crore in annual CSR disbursements is project-coded, sector-tagged, and mapped to a company’s net profit. The database is public, searchable, and updated each fiscal year.

Nobody uses it to decide where to spend.

Since Section 135 of the Companies Act came into force in 2014, India’s mandatory CSR spend has more than tripled. Healthcare and education together absorb 54% of combined disbursements. The MCA dashboard tracks every project, every rupee, every company — one of the most detailed social investment databases produced by any government anywhere.

This is a genuine achievement. Capital that was previously at the discretion of corporate boards is now obligatory, and the transparency is real. On the specific question of whether companies are spending their mandated 2% — India can answer it precisely, at scale, in real time.

The question it cannot answer is what happened next.

Maharashtra, Gujarat, and Karnataka together receive approximately 40% of India’s CSR spend. The eight northeastern states receive under 3%, combined.

Before treating this as a failure, the framework’s design logic needs to be stated plainly. Section 135 links CSR obligation to where a company operates and earns profit. A refinery in Surat faces community pressure in Surat. A technology campus in Bengaluru draws from labour markets and infrastructure that the surrounding district depends on. Social licence to operate is a local condition — and the expectation that CSR investment follows the business has a coherent rationale. It was a design choice, not an oversight.

Infrastructure, connectivity, and private capital have concentrated in western and southern India for decades. The Northeast’s low share of social investment is, in significant part, a function of its low share of industrial presence. Companies are, largely, operating within the framework’s intent.

What the framework was never designed to address is the gap between where India’s industries are and where India’s needs are most acute.

The chart below plots eight states across two axes: their share of national CSR spending, and their multidimensional poverty rate as measured by NITI Aayog’s MPI. Bubble size indicates industrial presence as a share of national GVA.

The pattern holds without exception across these eight states. Industrial presence predicts CSR spend. CSR spend has no relationship to need. The two lines of causation run perpendicular to each other, and the framework provides no bridge between them.

The intelligence deficit is not only geographic.

The MCA database records outputs: rupees spent, sector tagged, company named. It does not record outcomes. A company that spent ₹500 to improve one child’s reading level by one grade appears in the same database row as one that spent ₹5,000 for the same result. A school building is logged. Whether it has qualified teachers is not a required field.

CSR reports submit “10,000 saplings planted.” The survival rate — the figure that determines whether the spend produced any environmental impact — is not in the form. The BRSR framework, which applies to the top 1,000 listed companies, mandates ESG disclosure. It requires disclosure of what was done. Whether it worked is not in scope.

The consequence is that the biggest funding decisions — which NGO to back, which district to anchor a programme in — are still made through personal networks and press coverage. High-visibility, well-connected organisations attract capital regardless of demonstrated results. Organisations with verifiable outcomes, working in districts outside any CSR committee’s direct experience, remain invisible. The database exists. It contains no information that would change this.

India built a mandatory CSR reporting system detailed enough to be genuinely useful. It required spending. It required disclosure. It did not require impact measurement, and it did not build a mechanism for the case where industrial geography and need geography point in opposite directions.

Those are not the same as the system failing. They are the system reaching the edge of what it was designed to do — and the design stopping there.

The MCA dashboard will record over ₹35,000 crore in spending next year. It will not record how much of it moved any needle on poverty or public health. Both facts will be true simultaneously, and neither will appear on the same dashboard.

India built the most detailed CSR reporting system in the world and made it mandatory. It did not make intelligence mandatory. Those are not the same thing.