The Elephant and the Flood Are Now Insured Events. Proving the Loss Is the Work.
Two of the commonest ways an Indian crop is lost have sat outside crop insurance for years. An elephant herd walks through a paddy field two nights before harvest, or a coastal plot sits under water for a week, and the loss is real, local and — until this season — nobody's payable event. The Agriculture Ministry's modalities under the Pradhan Mantri Fasal Bima Yojana change that. Crop loss from wild animal attack is now the fifth Add-on Cover under PMFBY's Localised Risk category, and paddy inundation is back as a localised calamity. The rollout began with Kharif 2026 — the season whose harvest is now ending. So these covers are no longer an announcement. They are a claims pipeline that has just run its first season, and the interesting questions are all about proof.
What actually changed in the rulebook
Start with the two covers separately, because they fail differently. Wild animal attack — the ministry names elephants, wild boars, nilgai, deer and monkeys — was simply not a covered event. Losses went uncompensated, in the ministry's own words, however often they happened near forests, wildlife corridors and hill terrain. Paddy inundation is a return, not a debut. It was removed in 2018 because of moral hazard and the difficulty of assessing a submerged crop. That removal reason is the sentence underwriters should keep. The risk did not change in 2018 or since. What changed is the verification machinery the scheme is now willing to trust, and whether that machinery works is what the first seasons will decide.
| The two new localised covers | Wild animal attack | Paddy inundation |
|---|---|---|
| Status under PMFBY | Fifth Add-on Cover, Localised Risk category | Reintroduced as a localised calamity (removed in 2018) |
| How a loss enters the system | Reported within 72 hours on the Crop Insurance App, with geotagged photographs | Localised-calamity assessment in notified flood-prone units |
| Who defines exposure | States notify the animal list and identify vulnerable districts/insurance units from historical data | Coastal and flood-prone districts, per the ministry's state list |
| States named by the ministry | Odisha, Chhattisgarh, Jharkhand, Madhya Pradesh, Maharashtra, Karnataka, Kerala, Tamil Nadu, Uttarakhand, and the Himalayan and North-Eastern states | Odisha, Assam, West Bengal, Tamil Nadu, Kerala, Karnataka, Maharashtra, Uttarakhand |
| Rollout | From Kharif 2026 | From Kharif 2026 |
Source: Ministry of Agriculture & Farmers Welfare release on the PMFBY modalities (Release ID 2191224). Farmioc.
72 hours and a geotagged photograph
The reporting rule for animal damage is specific: the farmer reports within 72 hours, through the Crop Insurance App, uploading geotagged photographs. Read that as an insurer, not as a beneficiary. A localised cover lives or dies on the gap between the event and the evidence. A trampled field photographed three days later, with a location stamp, is assessable. The same field reported at harvest is an argument. The 72-hour window is therefore not administrative tidiness — it is the fraud control that replaces the crop-cutting experiment for this class of loss.
It also creates a data stream insurers have not had to process at scale before: high volumes of small, individual, photo-based claims, arriving within days of an event, concentrated in forest-fringe districts. The operational question for this rabi is unglamorous. Can district teams verify a geotagged trampling claim against sowing records and forest-fringe maps fast enough that settlement stays inside the season? If verification lags, a localised cover settles like the old wide-spread claims did — late, and into the next crop's borrowing.
The state list is now an underwriting document
One clause in the modalities deserves more attention than it has had. States will identify vulnerable districts or insurance units based on historical data, and notify which animals count. Two consequences follow. First, coverage quality will vary by state in a documented way: a district that does the historical work properly gets a defensible vulnerability map; a district that notifies late or thinly leaves its farmers with a cover on paper. Second, those notified lists are public underwriting information. A lender with crop-loan exposure in a notified elephant-corridor district is looking at a different risk than it was a year ago — not because the elephants changed, but because the loss is now partly insured, which changes the borrower's recovery after a bad night.
The same logic runs through the inundation list. Paddy in the coastal and flood-prone states named by the ministry — Odisha, Assam, West Bengal, Tamil Nadu, Kerala, Karnataka, Maharashtra, Uttarakhand — now carries a cover that was withdrawn eight years ago precisely because submerged crops were hard to assess honestly. Satellite water-spread data and the season's own flood record give assessors a cross-check that barely existed in 2018. Whether that cross-check is actually used, claim by claim, is the difference between a cover and a leakage channel.
The settlement machine behind the covers
None of this matters if the payment end stalls, and here the recent record is genuinely better than the scheme's reputation. On 24 October, Madhya Pradesh moved ₹317.86 crore of PMFBY claims for Rabi 2025-26 in a single-click transfer, covering around 16.70 lakh eligible applications, according to the scheme's official updates reported that day. The same updates put the state's cumulative claims since 2016-17 at over ₹32,400 crore through Kharif 2025. Scale is not the constraint any more. The constraint is whether localised, photo-evidenced claims can move through the same pipe at the same speed, or whether they queue behind the area-based claims they were designed to bypass.
What to do with this
For insurers: price the add-on cover off the notified district lists, not off state averages, and audit your first season of animal-attack claims for the obvious pattern — reports clustering at the 72-hour edge, or photographs that do not match the sown crop on record. For lenders in the named districts: re-mark which of your borrowers' worst-case losses are now insured events, and stop carrying the full trampling and submergence risk in your own provisions for the ones who are covered and enrolled. For both: watch whether states publish their animal lists and vulnerability maps before rabi enrolment closes. A cover whose map is late is a cover whose claims will be disputed, and disputed claims land back on the borrower's balance sheet — which is to say, on yours.
DataStore holds the district-level crop, weather and price series that sit underneath assessments like these, so an insurer or lender can check a localised claim against what the season actually did in that district, not just against the photograph that reported it.