But I think there are two fundamental problems, and on these issues the government of the day needs to bite the bullet. The first one is easier; what is its politics ? Does the government want crop insurance to be essentially a dole? A massive non actuarial social blanket that guarantees compensation on account of crop loss like the NAIS, that it can afford in perpetuity and intervene at will, without limits (which I think is politically expedient but financially unsustainable). Or it would like to engineer market based insurance products in partnership with insurance and re-insurance which limits its own liability to the premium ? A situation where the central government acts as a regulator encouraging transparency, vetting insurance products, influencing insurance companies and guiding state government; most importantly reigning in rogue behavior by state governments and insurance companies alike. (The latter path is politically difficult but better as policy predictability will reduce the cost of premium and cover more farmers).
When the agri insurance market started getting privatized in 2007 with the advent of WBCIS and later in 2010 with the advent of MNAIS, state governments and politicians started complaining that insurance companies would not pay out every time and “keep” the money. This is where the essence of insurance has to be bought. Any insurance company collects premium, builds a pool and uses that pool to pay when a trigger strikes. It cannot pay every time. Insurance is not dole, it never was and never will be. If the government can afford a non-actuarial dole programme like the NAIS where it is exposed to unlimited liability, then by all means. But if that is not an option, then as I said you need to bite the bullet.
The second problem is a harder one because it derails even an efficient dole programme. When policies are presented to farmers no policy is considered satisfactory because it does not deal with the problem of “basis risk”. This means that there not enough data to satisfy the farmer at the individual level. If there is a weather based product he/she wants data practically on his field and if there is a yield based product he wants it for his crop for his field. Insurance products have an area approach, and over the years governments have tried to increase resolution to village, block or tehsil from down from the district level. But to do this we all have to accept that state capacity to collect data especially yield and acreage is perfect and that the government with all its agencies can actually improve upon it. Since agriculture statistics are collected by the state, there is huge amount of variability in the quality of data.
Some states like Karnataka and Gujarat are meticulous and use technology but and in many the actual CCEs are barely conducted and the use of remote sensing technology is limited. There are instances where there is extortion by the lekhpal to where farmers and lekhpals ally each other to write below normal yields and share the payout. There is no comprehensive counter check or audit system. Any policy or compensation dole system will fail to take off until this problem is fixed. The solution lies in privatization of data collection. And weather is an instance where privatization has actually worked. the resolution for sampling weather data has increased to 5 to 15 km with help of private data providers.
The private network of weather stations numbers approximately 4000. This twice the size of the state owned network, and if we add the government network of about 1400 systems: jointly the density comes to on an average 9 observatories/district across India. This vast infrastructure expansion was achieved by 0 capital expenditure by the government. Now there is technology where this high density of automatic weather stations can be mixed with data from doppler weather radars to identify rain or the lack of it on the farmers field.
On the yield private companies have already used drones in India to collect acreage and yield data at the farmer level. The AICIL has already conducted a pilot with Gujarat. Rajasthan has conducted a pilot too. Yes there is today technology to identify every farmers field and see it with the naked eye. This is not science fiction, and it has been done in India. A combination of drones and high resolution satellite images is the answer. The solution has and will come about by the public, private partnership of technology companies, governments and insurance companies. The central government has to look outside its own system.
The government wants to introduce a farm insurance guarantee scheme (FIGS). From what I understand this is an index of weather and yield. The idea is excellent, because it will guarantee a base income of the farmer. But if this is based on the current yield, acreage and price institutional data collection system, it will have the same problem that yield based policies have. The government could look at price data from Reuter Market lite (RML). The RML has been collecting good clean price data from mandis all across country for six years or so (even horticultural crops).
Privatization of data collection will substantially reduce the cost of premiums, something that governments both and state and central desire.
This current government has shown resolve, it has stood by the land acquisition bill as well as GST. Both of which are politically difficult and require hard decisions. Crop insurance too needs to be dealt in the same way.