Recurring Floods, Funding Gaps: The Cycle Assam Remains Stuck, Year After Year

In 2024, Assam's flood damage was placed at ₹534.45 crore. Central assistance for that event came through the National Disaster Response Fund (NDRF), in two tranches of ₹1,270.78 crore and ₹313.69 crore. It was cleared only in October 2025, more than a year after the disaster itself. By the time that money arrived, the state had already been through another monsoon.

Between 2012 and 2020, Assam received no NDRF assistance at all, despite annual flooding.

Central releases to the North-Eastern states remain modest relative to disaster frequency. Parliamentary data show that Assam received ₹716 crore in State Disaster Response Fund (SDRF) support and just ₹29 crore in NDRF assistance in 2024-25. Manipur, Meghalaya, Mizoram, Nagaland, Sikkim, and Tripura together received well under ₹700 crore in combined SDRF releases the same year. This is despite a preferential 90:10 Centre-State funding ratio designed specifically for Northeastern states.

A region that floods every year is being financed as if each flood were a surprise.

Additional central assistance, beyond a state's own SDRF allocation, is released only after an Inter-Ministerial Central Team surveys the affected area and submits its findings. This process takes months by design. It is also meant to fund response and recovery, not to compensate for the full extent of loss.

How Much Of India's Disaster Spending Is Ex Ante/Ex Post?

In FY2025-26, the Centre disbursed ₹13,603 crore under SDRF and ₹2,024 crore under NDRF. Both are ex post (spent on relief and reconstruction after the event) relief instruments. Only ₹4,571 crore was released from the State Disaster Mitigation Fund in the same year, the newer instrument meant to fund ex ante (before the event) preparedness.

Mitigation funds created under the 15th Finance Commission have gone largely unused, even as relief spending keeps rising. Infrastructure carries the cost of this imbalance. A Coalition for Disaster Resilient Infrastructure study estimates that 93% of India's infrastructure disaster risk remains uninsured. The same study puts India's cumulative infrastructure losses to weather damage between US$619 billion (nearly ₹60 lakh crore) and US$1.02 trillion (nearly ₹100 lakh crore). A separate asset-exposure study for Odisha found ₹1.81 lakh crore of public infrastructure at risk, across ten asset classes including highways, irrigation works, schools and hydropower assets.

This uninsured exposure falls on the government almost entirely today: India's average annual disaster loss is close to US$9 billion (nearly ₹87,000 crore). Less than 2% of this is covered by catastrophe insurance, well below the global non-life insurance average.

Closing even part of that gap would mean that a significant share shifts from government budgets to insurance and reinsurance markets.

Still, it raises a pertinent question: why should the exchequer keep underwriting a risk that markets are willing to price?

Decentralisation Needs Decisions

Closing that gap is not only about buying more insurance. It is about who gets to make decisions, and how fast, when a disaster hits.

A policy brief by the Policy Consensus Centre, Evaluating the Efficacy of Decentralised Disaster Risk Financing in India (June 2025), found that funds must move together with decision-making power, as decentralising finances alone accomplished marginal results. Funds must move together with decision-making power. Odisha's post-1999 model makes this case well. It devolved discretionary authority over food, transport and evacuation decisions to district collectors and block-level officers. The result was measurable. Cyclone Dana, a storm of comparable intensity to earlier disasters, caused zero recorded loss of life in Odisha in 2024. Nagaland's parametric insurance pilot offers the financing-side parallel. Its first iteration, running from 2021 to 2023, failed to trigger a payout because its rainfall dataset diverged from India Meteorological Department (IMD) readings. Nagaland corrected this flaw in a revised three-year cover worth ₹150 crore, using locally calibrated triggers, running from 2024 to 2027.

The Disaster Management (Amendment) Act, 2025, has been in force since April. It offers a legislative opening to generalise this shift. States can now constitute Urban Disaster Management Authorities with dedicated resilience funds. The Act also mandates a national disaster database and expands the decision-making role of District Disaster Management Authorities. In practice, adoption has been slow. As of early 2026, only Karnataka has constituted an Urban Disaster Management Authority (for Bengaluru's municipal corporation). A law that permits decentralisation is not the same as a state that practises it, and this gap is itself a measure of how much work remains.

Assam is the immediate case in point, but these steps are not specific to one state or one kind of disaster. The same five measures apply wherever local risk is high and central response is slow.

Decentralised Measures

Firstly, pilot village-level, rainfall-indexed insurance in Assam's most flood-affected districts. It should be calibrated against IMD data, not third-party datasets. This directly addresses the trigger-mismatch failure Nagaland encountered in its first attempt.

Secondly, use the amended Act's provisions to devolve real decision-making authority, not just funds, to District Disaster Management Authorities. Follow Odisha's model of pre-disaster discretionary power at the district and block level, as it determines whether decentralised financing actually shortens response times.

Thirdly, operationalise the mandated national disaster database with granular, real-time, state-wise data on loss and expenditure. This would end relief's current dependence on a lagging post-disaster survey.

Fourthly, revisit SDRF and NDRF allocation formulas for the Northeastern states against actual disaster frequency, rather than population or area alone. This would close the gap between exposure and entitlement.

Fifthly, and most structurally, explore a dedicated, SPV-based decentralised disaster fund open to public and private capital, including insurance and reinsurance participation. This would shrink both the multi-year lag between loss and disbursed relief, and the share of that loss the treasury absorbs alone.

None of this requires waiting for the next flood to test it. The instruments already exist: a functioning Act, a template in Odisha, a correctable failure in Nagaland. What is required now is the will to connect them, and to build the kind of climate-resilient governance that Assam, and every flood-prone state after it, actually needs.

(The writer is the Co-Founder & CEO of Policy Consensus Centre. Views expressed are personal.)