Disaster finance cannot begin only after a disaster. States need to know in advance which losses they can absorb, which to insure and those that must be shared more widely.
Disasters like the one that struck Nepal expose an important weakness in how countries like India financially prepare for extreme events. While investments in disaster response, mitigation and reconstruction are essential, the financing of residual losses often remains largely reactive. In many cases, the question of who ultimately bears the economic cost—governments, households, businesses or insurers—is resolved only after the disaster has occurred. India needs to change that.
When a flood destroys a small enterprise, a heatwave reduces a worker’s earnings or a cyclone damages public infrastructure, the economic loss ultimately lands somewhere. A household runs down its savings. A business loses revenue. A borrower struggles to repay a loan. A state government finds additional money for relief and reconstruction.
Uninsured risk is, therefore, not risk that has disappeared. It is risk that governments, households and businesses have retained often without consciously deciding to do so.
This matters because India’s economic exposure is rising. Estimates suggest that $26-29 trillion (around Rs 2,500 lakh crore) of assets across the country are exposed to different natural hazards. Across Asia, only about 8 percent of economic losses from natural catastrophes in 2025 were insured. The answer is not to insure everything. Rather, India needs to decide systematically which risks should be borne by whom.
A useful starting point would be for every state to prepare a disaster-risk financing strategy alongside its disaster-management plan.
States already assess hazards and vulnerability. They should now estimate their potential financial exposure: what could a severe flood cost? How much loss could existing disaster funds absorb? At what point would reconstruction begin crowding out other development expenditure? Which risks are predictable enough to insure?
This would create a simple hierarchy. Frequent, smaller losses could continue to be met through public funds and existing safety nets. Larger but measurable shocks could be partly insured. Truly catastrophic events could be shared through reinsurance or other mechanisms so that neither a single insurer nor a state budget bears the entire loss.
The timing is particularly relevant. The Sixteenth Finance Commission has recommended Rs 2,04,401 crore for state disaster response funds between 2026-27 and 2030-31, including Rs 40,880 crore for mitigation. It has also asked the government to examine and finalise single-peril parametric insurance products.
Some states have already begun experimenting. Nagaland has purchased multi-year parametric cover against excess rainfall. Unlike conventional insurance, where losses are assessed after an event, parametric insurance releases a predetermined payment when an agreed threshold such as rainfall, temperature or wind speed is breached. Existing disaster-mitigation guidelines have also been interpreted as permitting states to use part of these funds towards insurance premiums.
That creates an opportunity, but it also requires guardrails.
Public money should not simply become a new source of insurance premiums. States should first demonstrate that transferring a particular risk is cheaper or more effective than retaining it. Premiums should be competitively procured, and governments should periodically evaluate whether the cover is actually reaching the losses it was intended to address.
The second challenge is trust.
Parametric insurance is fast precisely because it does not measure each person’s actual loss. But this also creates the possibility that people suffer damage while the official trigger is not crossed. In heat-insurance experiments, comparisons of weather datasets have shown differences of around 3-3.5°C, enough in some places to determine whether a payout occurs.
India therefore needs a public framework establishing which weather data can determine payouts, how thresholds are validated and what happens when data sources disagree. The India Meteorological Department, disaster-management authorities, insurers and state governments will need a common protocol. Without that, scaling parametric insurance could create distrust precisely among households that have the least capacity to absorb a failed payout.
There is also a case for states to cooperate rather than buy disaster cover entirely on their own. India’s climate risks vary enormously across its geography. A cyclone affecting the eastern coast may have little relationship with drought conditions in western India or excess rainfall in the Northeast. Pooling carefully selected, imperfectly correlated risks could potentially allow states to share administrative costs and negotiate more efficiently with insurers and global reinsurers.
But financial protection cannot replace physical protection. Better drainage, early-warning systems, heat-action plans, resilient roads and safer construction must remain the first line of defence. Insurance should finance residual risk, not become an excuse to tolerate avoidable vulnerability.
That distinction is particularly important for vulnerable households and informal workers. The State cannot withdraw from its responsibility for disaster relief merely because a financial product exists.
The important question before the next flood, cyclone or heatwave is not simply how much money will be available afterwards. It is whether governments have decided beforehand which losses they can absorb, which they should insure and which risks must be shared beyond the budget.
Ria Sinha
Senior Fellow at Chintan Research Foundation and former Senior Fellow at the Indian School of Development Management
(Views are personal)
