This post is part of our Closing the Gap: Financial Inclusion  blog series, which shares the views of selected experts and practitioners on different financial inclusion topics.
How do you insure hundreds of millions of small farmers spread over many developing countries? There is no easy answer. Individual insurance would entail assessing crop yields in millions of farms within the short harvest windows – a virtually impossible task. And even if this were possible, costs would be prohibitive and data quality, a significant issue.
Yet the importance of finding a solution cannot be underestimated. First, farmers make up the majority of the world’s poor. With high dependence on rain-fed cultivation, agriculture is risky. Mitigation of those risks is critical to stabilizing the income of poor farmers. Otherwise, a crop failure could erode savings, lead to inability to service crop loans, push farmers into a vicious debt trap as they are forced to borrow from moneylenders and in extreme cases, lead to starvation or even worse.
Second, crop insurance can in effect act as collateral against loans and enhance the viability of crop lending – hence it is vital for banks. Effective crop insurance could potentially be one means to address the confounding conundrum in agriculture credit, namely, how to undertake viable lending to a high risk activity, especially given that the political economy and farmer level economics make market-based pricing of insurance unfeasible. If the price of risk mitigation increases, political economy and farm-level economics get disrupted and credit risks are raised. If prices stay level , viability concerns lead to credit rationing. Crop insurance is a possible way out of this catch-22 situation.
Third, increasing the creditworthiness of farmers can in turn offer them an opportunity to invest in appropriate inputs to increase agricultural productivity, thereby contributing to overall sector productivity.
Thus, an effective crop insurance program is critical for overall agriculture development which impacts not just risk mitigation for farmers, but also agriculture credit and agriculture productivity.
That brings us back to the question we started with. How does one insure hundreds of millions of farmers with small individual landholdings?
While crop insurance has existed for many years, practitioners and policy makers have recently found a way, with support from the World Bank and others, to use new tools to index, claim and compensate farmers. Since individual insurance is not feasible, ‘indexed insurance’ has been applied. The idea is simple. The program defines a proxy index for crop production and rules for a common claims payment for all farmers in a given area. This mitigates moral hazard and adverse selection risks inherent in insurance.
Two types of indices have been applied at scale. These include the traditional ‘area yield index’, and ‘weather indexed’ insurance, which have both been scaled up in recent years. In the former, if the average of sample yield measurements in a given geographical area is lower than historical averages, a claim is payable. In the latter, weather data triggers a payout. Triggers include proxies for crop yield, such as temperature or rainfall relative to a particular threshold in different phases of the crop cycle.
Though both approaches represent innovative thinking, increasingly weather indexed insurance is being talked about as the panacea and as a substitute for ‘area yield insurance’. For area yield insurance, data gathering remains a logistical challenge, and poor data quality and delays have been common. For example, it is not uncommon to hear about agriculture field officers ‘filling up’ forms on sample crop yields while sitting at home, or underreporting yields in response to local political pressure.
Indeed, weather indexed insurance is an innovation that needs to be celebrated. Most importantly, it offers independent and easily and quickly measurable data that can offer quick settlements of claims, thereby benefiting farmers. For these reasons, private sector insurers are coming in in a big way, enhancing competition and product choice.
But as with many other things, the verdict needs to be more nuanced. One, the quality of correlation of weather indices with crop production may vary. Just to cite two instances: In Rajasthan, a state in western India, Rabi (winter) crop farmers were delighted with a record payment despite 2010 being a record crop production year. But in 2009, in the same state, low payments were made in what was clearly a bad crop production year. Second, weather insurance offers single peril insurance, leaving out protection against, say, a pest attack. Third, product design and data availability pose a challenge. Fourth, low density of weather stations means that local weather events do not get captured, and like in ‘area-yield’ insurance, basis risk (mismatch between farmer yield and area index) exists. Fifth, the independence of weather data is now being put to question. Anecdotes show that some farmers are now gaming-the-system. For instance, it is easy enough to pour some water into the weather station to trigger ‘excess rainfall’. Or drop ice into it, to trigger a temperature claim!
So does index insurance really provide the answer to our initial question? Our work in the World Bank shows that yes, it does.
Despite imperfections, indexed insurance is a great innovation. With more research and product development and some checks and balances in implementation, weather insurance can become even better. But like the case of microfinance being sold as a silver bullet for poverty alleviation, let us not make the same mistake here (and likewise, as with microfinance, let us not throw the baby out with the bathwater either). And despite imperfections, area index insurance can have value, as it is possible to address its deficiencies. Our current engagement tries to address its data quality and data speed issues through use of mobile (GPS, GPRS, camera) and remote sensing technology, while also improving and modernizing the design and delivery of both weather and area yield index insurance.
Indeed, our broad conclusion for India is that combining the good features of both approaches to indexed insurance could offer the best of both worlds. India is trying this out and if this succeeds, the relevance could be global.