Index-based Insurance and Climate Information in Sri Lanka: Prospects for Community-based Rainfall Stations

Kanchana Wickramasinghe

Climate insurance for farmers has been a hot topic recently owing to recurring floods and droughts in agricultural districts in Sri Lanka. In these instances, interventions are necessary to help farmers adapt to weather and climate changes, using effective risk management strategies. Global experience has shown that insurance serves as an important risk management tool in this regard. However, an IPS study indicated that, in Sri Lanka, climate insurance is not yet considered a popular risk management strategy adopted by farmers due to issues in supply and demand. 

Supply-side issues are related to the indemnity-based crop insurance products offered in Sri Lanka. The study proposed to shift to index-based insurance, to eliminate problems of traditional indemnity-based crop insurance. It further identified the lack of climate data as a main barrier in implementing index-based insurance in Sri Lanka.  

Index-based Insurance and Climate Information 

Once designed, the successful implementation of index-based insurance requires timely and accurate data, with precise geographical representation. This is especially important as insurance pay-outs in such schemes are purely based on data produced by the weather stations. An insurance payment is made only if the value of the index is above or below pre-determined threshold levels. However, a major challenge is to overcome the ‘basis risk’, the possible deviation between rainfall experienced by individual farmers and rainfall recorded at the stations. As a result, index measurements may not always tally with the actual losses incurred by individual farmers. Though basis risk cannot be completely eliminated, it should be minimised through data availability, to avoid possible issues in upscaling. 

As insurance products are designed and implemented based on selected weather parameters, index-based insurance requires climate data. The weather parameter has to correlate closely with crop yield. Rainfall is the most suitable parameter for developing index-based insurance in Sri Lanka. As highlighted by previous studies, developing a rainfall index-based insurance product requires, at the very least, historical daily rainfall amounts at each rainfall station.

A private insurance company in Sri Lanka has already made a preliminary attempt to design and implement index-based insurance based on rainfall. Unfortunately, data availability was a main hindrance in replicating the said programme. In Sri Lanka, the Department of Meteorology, by its mandate, undertakes the measuring of rainfall at regional rainfall stations. In addition, there are a few other government agencies that collect rainfall data on a limited scale, such as the Department of Agriculture, the Irrigation Department, the Department of Agrarian Development, and the National Building Research Organisation.  However, the current weather station facilities in Sri Lanka are not adequate to successfully implement a fully-fledged index-based insurance scheme.

Recent developments in Sri Lanka point towards improving the available climate information. The Budget Proposals for 2018  made a significant allocation to advance the technical capacity of the Department of Meteorology. This can enhance the accuracy and the timeliness of the available data. Moreover, the budget has also proposed to introduce weather-based index-insurance products as a risk management tool for farmers.

An alternative to existing rainfall stations is to establish automated rainfall stations. A private insurance company is currently attempting to establish automated rainfall stations to remove data issues that are hindering the implementation of its own index-insurance products. Ideally, data collected through such initiatives should be shared among other users too.

Community-based Rainfall Stations: Way Forward

Basis risk can occur due to snags in product design or geographical factors. The latter is linked to data availability issues. Basis risk tends to be higher when microclimatic variations in a given geographical area are high, as these variations may not always be captured by weather data collected in a station. When the distance from the weather station to individual farmers is high, mismatches are more likely to occur between actual rainfall and recorded rainfall. When the distance is higher, it can also lead to mismatches between actual losses and losses indicated in the index. This can mean that affected famers do not receive insurance payments. However, high density weather stations can minimise this risk.

Community-based rainfall stations are an innovative way to address this problem, while overcoming resource constraints associated with government agencies in generating data. The pilot-scale action research study, being carried out by the Institute of Policy Studies of Sri Lanka (IPS), in collaboration with the Department of Meteorology, and Janathakshan show that there is potential for a community-based approach in Sri Lanka. The project intends to bridge the climate information and communication gaps among the farming communities in Sri Lanka. An important element of the project is to focus on providing necessary equipment and training to farmers and getting them involved in measuring rainfall, using a standard approach. Farmers show a high level of enthusiasm in this regard. Daily rainfall data recorded by farmers is now shared with the Department of Meteorology on a periodic basis. Even though the pilot study introduced community-based weather stations as a tool for enhancing farmers’ adaptation decisions, it shows its potential to fill the current data gaps in index-based insurance schemes too. 

Despite signs of improvement, however, further developments are necessary to generate and share climate data, before successfully implementing weather–based index insurance schemes in Sri Lanka. It is necessary to ensure the reliability and validity of the data generated through community-based stations. Data should be controlled for quality and reported based on standards. The aim should be to develop an effective system where data is collected and shared through an optimal data management system. If community-based weather stations are to lead to the development of successful index-based insurance products, the data generated by these stations should be acceptable to both public and private insurance providers. To this end, it is useful to develop public-private-community partnership models for managing community-based weather stations.

(Kanchana Wickramasinghe is a Research Economist at the Institute of Policy Studies of Sri Lanka (IPS).