This brings to a crescendo the rising chorus for farmer compensation. Another farmer, Ram Sahey, from Dhanawa, near Sultanpur, in Uttar Pradesh, has lost 50% of his wheat harvest to hail, losing over Rs 2 lakh of potential income. With an additional debt of Rs 3.5 lakh, he expects compensation to be around Rs 100 and Rs 5,000. This is farming in India today.

Unseasonal rainfall during February and March damaged crops over 8.5 million hectares across 14 states, including UP, Haryana, Madhya Pradesh and Punjab — about one-sixth of the Rabi sowing area. About 3.9 million hectares of wheat were potentially damaged, with yield loss reported at 15%.

Central and state governments are taking mitigation measures, with compensation for crop damage raised by 50%, while eligibility norms are lowered to 33% crop damage. Insurance firms and banks have been pushed to play a proactive role through pay-outs and restructured loans. Faster compensation identification and dispersal, along with expanding insurance coverage, remain key. Insulting farmers with a pittance of compensation will not do.
This vulnerability is a historical actuality. Even a modest decrease of 10% of mean rainfall can lead to a significant decrease in rice production. The potential for future seasonal variations remains. During the period 1877-2009, India experienced 24 major droughts. During 2009, a rainfall deficit of 23% was recorded and about 59% of the area was affected.

Recording losses quickly and accurately is necessary. Belgaum district, in Karnataka, carries out data generation through field information facilitators — an example of decentralisation of responsibility and local capacity building. The utilisation of open software, like Google Earth, to speed up compensation claims could be another avenue.

In India, weather forecasts are valid only up to 4-5 days, while other extra-tropical regions have forecasts till two weeks. Studies on the impact of day-to-day variations of rainfall, particularly close to the harvest stages, have been mostly neglected. The accuracy of rainfall prediction in India has been error-prone.

An estimate of 20% deficit rainfall nationally means little for a local farmer, given wide district-level variation. Blanket agro-met district-level advisories need to be converted into block- and village-level ones.
Weather-based crop insurance schemes (WBCIS) have been introduced, with the National Agriculture Insurance Scheme (NAIS) insuring over 25 million farmers.

However, challenges remain, with 95 million remaining uninsured and just 6% of farmers voluntarily purchasing cover. Assistance and insurance payouts for crop damage are often delayed by a year, subject to the whims of local officers. Data quality too varies considerably between states — given differing levels of accountability, capacity and expertise of relevant agencies.

Worldwide experience suggests agricultural insurance is most efficient and effective when the private sector is encouraged to participate. Sampling error should be reduced by conducting sufficient crop-cutting experiments per insurance unit, while the area yield index insurance should accurately reflect aggregate shocks. The insurance unit should be reduced from the level of a block to a village, as recently carried out in Telangana.

Risk classification is also variable with the WBCIS and the NAIS rules for designing and pricing products, leading to the coverage value varying significantly across regions and years. Poor risk classification leads to significant adverse selection, with the farmers voluntarily purchasing cover in high-risk insurance units while promoting inequitable distribution of public subsidies and providing poor agriculture policy signalling.

Basis risk too remains considerable. A farmer can experience a large crop loss but receive no claim payments because the average yield in the insurance unit is not low enough to trigger a payment. Poor and risk-averse farmers are hindered, as a purchase worsens the worst that could happen.

Weather and area yield indices need to be combined, combining the strengths of the modified NAIS and the WBCIS, leading to more accurate loss estimates and faster claim settlement. Risk classification should provide flexibility to determine premium rates and threshold yields on an actuarial basis, instead of inflexible rates determined by simple formulae,leaving significant inequity across regions.

Basis risk can be reduced by combining farmers into local and semiformal mutual institutions.
The government should reconsider the overall agricultural risk market infrastructure, while assessing the merits of alternative models. For example, Mexico utilises a public reinsurance company to offer technical assistance and reinsurance capacity to the domestic insurance companies involved in agricultural insurance. Spain deploys a lead insurer as part of an agricultural coinsurance pool.

A shift to a timely and affordable market-based crop insurance programme with actuarially sound premium rates can shield farmers from the vagaries of unseasonal rainfall.

It can lead to significant benefits for farmers, including faster claims settlement, a more equitable allocation of subsidies and a lower basis risk — all the while lowering adverse selection,improving agricultural policy signalling and reducing dependency on government bailouts. Such large-scale reforms could alleviate farmer distress in the short term.