This post was originally produced during the GIIF Conference held in Paris at the OECD Headquarters.

The 6 key lessons learnt that emerge from the panel are as follows:

1. Insurance needs to be part of a value add package that generates higher income for the farmers
2. Governments play a key role in creating markets for agricultural insurance.  Experiences in Peru, India and Senegal point to government interventions such as premium support, risk financing, and linkage to credit.
3. There is a need to design and offer the right insurance product.  Area Yield Index Insurance may offer an adequate protection to farmers as it captures a wider range of perils than pure Weather Index insurance
4. Data is critical but it is not necessarily a precondition for insurance projects. Experience from Peru shows that data availability and reliability can be improved throughout the course of a project. 
5. Insurance may not always be the solution for the problem of climate change, and should be explored as part of a broad risk management framework. Agriculture insurance can help the transition to climate smart agriculture by securing investments in adaptation. In addition, by quantifying and mapping risks, insurance can also act as a catalyst for adaptation investments.
6. When it comes to index insurance, Knowledge often flows from South to North

During the GIIF index insurance conference in Paris, a panel of experts moderated by Panos Varangis,  Global Lead for Agriculture Finance at the World Bank Group, and consisting of Pape Ndiaye (CEO, CNAAS Senegal), Bhargav Dasgupta (MD & CEO, ICICI Lombard General Insurance), Christina Ulardic (Head of the Market Development Africa, Swiss Re Corporate Solutions),Rose Goslinga (Insurance Sales, PULA), and Lourdes del Carpio (General Manager of agricultural insurance, La Positiva insurance company) spoke on the factors that enable scaling up insurance and building a market.

This is with this image that Rose Goslinga, one the pioneers of index insurance, introduces the two expected benefits of agriculture insurance: on the one hand, insurance can protect farmers in bad years (the “lock”), and on the other hand, insurance can foster productive investments (the “flat screen TV”) through enhanced access to credit and incentives to invest. Recent studies have shown that access to insurance can act as a catalyst to boost farmers’ income and, according to Rose Goslinga, this is what farmers need the most: access to credit, access to quality and timely inputs, and access to markets. Christina Ulardic, Head of market development Africa at Swiss re, one of the largest global reinsurers, agrees: “Insurance alone doesn’t do it”. She highlights the need to leverage key stakeholders along the agriculture value chain - such as financial institutions, but more particularly agribusinesses – to sell insurance as part of a productive package. However, even when bundled with other value-added products and services, and with the exception of a few established programs (in India or Mexico for instance) most agriculture insurance programs reach only a few thousands farmers.

So what can Governments do to support sustainable scale-up of insurance markets?

Experiences from Senegal, India and Peru show that there is a wide variety of Government interventions to support agriculture insurance. In Senegal, Pape Ndiaye – CEO of agriculture insurer Compagnie nationale d’assurance agricole du Senegal (CNAAS) indicates – that Government support has been crucial on three fronts: support to the creation of CNAAS (35% of CNAAS is owned by the Government), 50% premiums subsidies and insurance tax exemptions. But as most of the 500,000 Senegalese farmers are still uninsured, he indicates that “more remains to be done”. In particular, he insists on the need to create insurance awareness, linkages to credit, data infrastructure, and enhanced insurance and reinsurance capacity. India is the birthplace of index insurance with the first pilot launched by ICICI Lombard in 2003 and is now home to the biggest agriculture insurance market with 26 million farmers insured. Such “scale-up” has been possible thanks to strong Government support both on the demand side (e.g. premiums subsidies, mandatory bundle of insurance with rural credit) and the supply side (e.g. investments in data collection). But again, more remains to be done as looking in relative terms, this accounts for 20% of farming households. In Peru, the Government has opted for a segmented approach: on the one hand, “catastrophic insurance” is highly subsidized and acts as a safety net for vulnerable farmers, and on the other hand un-subsidized insurance is sold through microfinance institutions (MFIs) to commercial farmers. Farmers covered under the subsidized scheme are 10 times as numerous as under the unsubsidized one. “For farmers it was a drought, but for the insurer it was just a dry spell. I wanted to give my own savings, or hide under a stone”. Rose Goslinga is quite open about sharing past failures with weather index insurance, in particular as regards to the key issue of “basis risk” – or the risk that a farmer experiences agriculture losses on his plot but does not receive insurance payouts. There is still no precise and widely accepted definition of basis risk but its causes can be summarized as follows:  (1) idiosyncratic risks faced by an individual farmer, (2) errors in data measurement both for weather data (due to low weather station density) or yield data (due to inadequate estimation procedures), (3) errors in weather index modelling, (4) or simply the fact that a farmer is exposed to perils which he is not insured for (e.g. weather index insurance can cover rainfall deficit or excess, but not pests and diseases). While Area-Yield Index Insurance products cover a wide range of perils (as they cover all types of risks that can impact the average yield in a defined area), Weather Index Insurance products are based  on weather variables are therefore more vulnerable to basis risk (the last two sources of basis risk are only present with weather index insurance). This is why Rose Goslinga believes that “Area-Yield Index Insurance is where the future lies: it is tough to do, but easy to explain to farmers as it captures all risks”. Pape Ndiaye also shares that after several Weather Index Insurance pilots, he is thinking about launching in an Area-Yield Index Insurance product. “We have offered rainfall deficit products, but this year our clients have mostly suffered from rainfall excess, so we need to think about a product that is better adapted to their needs’. Christina Ulardic disagrees: “I don’t think we should really frame the question as Area-Yield Index Insurance vs Weather Index Insurance but rather what product can be offered given the type of data which is available”.  But is data availability the real constraint?

Lots of people decide not to launch insurance products, because data is too scarce. Well, we decided to start, create data, and learn

Indicates Lourdes del Carpio. Data is considered as the “lifeblood of insurance”, as long and reliable series of yield or weather data at a granular level (e.g. village of group of farmer) are generally necessary for insurers to price products. Such data was not available 6 years ago, but her company decided to go ahead and launched a product with premiums amounting to 15%. Today, La Positiva has a much better understanding of risks, and premiums are now down to 10.9%. Rose Goslinga also highlights that “You will not find all the data in one place so you need to be creative in finding data and putting together these datasets”. Indeed a wide variety of stakeholders collect data both in the public sector (e.g. Ministry of Agriculture, Ministry of Health) and in the private sector (e.g. farmers’cooperatives, large buyers). However from the perspective of a reinsurer, it is quite important to be sure that such datasets are accurate and reliable before underwriting products. “We need to be sure that yield data is sampled, collected, and transferred in a reliable manner across time and locations” reminds Christina Ulardic, sharing that when switching to a new data provider, Swiss Re obtained data which was not consistent with previous data. Inconsistent and unreliable data is exactly what insurers and reinsurers are worried about, which is then translated into “ambiguity loads”, and therefore higher premiums for farmers. Beyond data availability and reliability, another key issue from a farmer’s perspective is the speed of the data collection process. Bhargav Dasgupta indicates for instance that when the National Agriculture Insurance Scheme started in India in 1999, crop cutting experiments undertaken by Government workers to assess losses were a very lengthy process, leading to claim payouts disbursed up to 2-3 years after production shocks. Much progress has been made since, and ICICI is now thinking about using new sources of data such as satellites or drones to complement data collected on the ground. There are indeed several innovations focused on combining granular ground-level data with generally inexpensive and large scale remote sensing data (e.g. rationalization of crop cutting experiments, or double-trigger products).;

I am afraid that we are speaking only of risk transfer, when we should speak of all risk management needs

Asks a voicefrom the audience. Insurance is indeed a one risk management solution among others, particularly suited to high-severity and low-frequency risks. As an insurer, Lourdes del Carpio is definitely interested in investing in risk prevention and indicates that as soon as El Nino was forecast earlier this year, La Positiva has launched with the Peruvian Ministry of Agriculture a radio campaign to inform farmers about “simple things that farmers can do” to mitigate its impact. Beyond individual level mitigation measures, insurance can also act as a “catalyst for other risk mitigation investments and adaptation to climate change at national level” indicates Panos Varangis. Indeed, one of the merit of insurance is to map and quantify risks, which can lead to better awareness of risk exposures and more investments in risk mitigation.

When it comes to index insurance, “Knowledge often flows from South to North

Highlights Cristina Ulardic. Indeed, earlier in the morning, Jean-Laurent Granier from AXA was indicating that an index insurance product was soon to be launched in France for farmers growing forage. The product will be based on satellite data and sophisticated modelling, but its overall principle is borrowed from about a decade of experience in developing countries. Similarly, some of the most advanced farming practices are actually found today in South Africa. Knowledge comes from unexpected places, and this is how participants leave this plenary session: ready to learn from the next speaker or the next coffee break encounter about this little thing – about data, basis risk, distribution models, Government incentives or risk management - which will help them achieve their goal. About 200 professionals from various countries and horizons -  insurers, scientists, donors, bankers, brokers - will share ideas this week so that, together, they can work towards better access to adequate and affordable insurance solutions for millions of farmers.