Providing Weather Forecasting Services along with Agricultural Insurance to Small Farmers in West Bengal Presentation on findings of Project End Report Kinnary R. Desai CIRM Why Agro-Insurance? Agriculture is important: -provides employment to 2/3rd of our population -growth and development taking place But, vulnerable: -affected by volatility in input and output prices, -credit availability, -government regulations, subsidies, support prices - storage and processing risks and - production risks like unfavorable weather, catastrophe, diseases and pests Agro-insurance: one of the market based formal methods risk management in agriculture;Weather insurance to manage weather risks Weather Index Insurance (WII) What is WII? “A product whose pay-outs are linked to a publicly observable index, such as rainfall recorded on a local rain gauge” - Gine,Townsend &Vickery (2008) Other weather indices: temperature, humidity, wind velocity, sunshine hours, etc. Advantages: -transparent -payout time minimized -adverse selection and moral hazard problems minimized -applicable in case historical yield data is absent Disadvantages: -high basis risk -complex framework, difficult to understand for farmers The Intervention Background: -Comprehensive Agriculture Risk Management Services (CARM) launched in six remote districts of India: Howrah, Hooghly, Bankura,East and West Medinipur in West Bengal and Kamrup in Assam; analysis based on Howrah only. -Partners:WRMS and CIRM -37 Automated Weather Stations (AWS) installed inWB and 15 in Kamrup to record rainfall. Why? The Product: -WII (rainfall index) for paddy and potato farmers ; policy underwritten by ICICI Lombard and AIC. -perils covered: excess rainfall for Kharif crop (paddy) and unseasonal rainfall and extremely high temperatures during the cropping season, resulting in continuous crop disease conditions for Rabi crops (paddy and potato) -WII was bundled with weather forecasts (24 to 48 hrs from time of delivery) broadcasted through mobile SMSs as a value add. Why? -Delivery channel set up by WRMS- rural banks, NGOs, microfinance institutions, cooperatives, and contract farming and input firms Research Questions and Methodology What are the factors that determine the take-up of the weather index insurance product by farmers? -Logit regression Model Has the dissemination of weather forecasts through SMS helped the insured to mitigate and manage risk posed by unfavourable weather? -Self Reported evidence of client farmers What is the impact of weather insurance on the production related decisions of the client group? -Feedback of client farmers Determinants of Take-Up Educational status/ product understanding & Risk Occupation & alternative means of risk management Client Characteristics • Education: school enrollment • Participation in Govt. Schemes • Agriculture as Main Occupation • Agriculture as Secondary Occupation • Off Farm Activities (X11 =1 if yes) Liquidity constraints • Logarithm of Annual Income • Jewellery and livestock Value added services • Number of Mobile Phones Response to agricultural shocks/risk perception Risk Covered Demographic Characteristics • Type of house: Kutchcha • Inputs financed through borrowing: paddy and potato Take-up • Possession of water pump • Weather forecasts important • Risk lover and among top 3 factors • Time to return to normal cons. • important for good yield • Area Cropped • Total Input Cost • Area owned (Bighas) • Weather crop shock loss • Joint Family • Age • Gender • Household Size • Category (general/SC/ST) The Sample Data collection: -1/3rd of the households (from 25 villages) from MIS information were listed between Dec. 2011 and Jan. 2012 -sample selection was not stratified, but, was randomized; challenges faced -final sample of 211 client and 211 non-client households was selected Descriptive statistics: Desc. stat. Value Desc. stat. Value Male 84.6% Hindu 98.1% Avg. age 43.5 yrs General Cat. 75.8% Avg. hhld. size 4.8 members Past schooling 91.7% Married 87.9% Main occ.- agri 83.4% Avg. area cropped 2.7 bighas Avg. annual inc. Rs. 37,745 Model Specification Logit Regression Take-up = β0 + 26 =1 βi Xi + Ɛ …………. (1) Results: Determinants of Take-Up Education: School enrollment has a significant positive impact on take-up; consistent with expectations 60.00% 50.00% 40.00% 30.00% Clients 20.00% Non Clients 10.00% 0.00% Didn't go Primary Secondary Higher to school education Education Education Occupation: -clients are less likely to have agriculture as main occupation -counter-intuitive; might be due to liquidity constraints Agriculture as occupation Main occupation Assets Type of household Clients Non clients Total 73.90% 92.90% 83.40% Type of household Clients Non clients Total Total Value of livestock (Rs.) 527689.9 343655.7 871346 Avg. value of jewellery (Rs.) 10349.3 6876.5 8612.9 Contd. Borrowing: - % of clients (86% paddy; 5% potato) financing inputs through borrowing is significantly higher than non-clients (68%; 1% resp.) -significant coefficient -take-up to repay loan to maintain relation with moneylender Mobile Ownership: - 61% of clients and 49% of non-clients own mobile phones -number of mobile phones positively correlated to take-up -access to mobile phone necessary for utilization of VAS Risk Profile: -risk lovers have significantly low take-up -consistent with a-priory expectations Payouts in each case (in Rs.) % of clients who prefer Option each option Heads Tails 1 50 50 42% 2 40 120 53% 3 0 200 5% Contd. Perceived importance of weather forecast: -70% of clients and 57% of non-clients ranked weather forecast among the top three requirements for good yield -significant positive coefficient Participation in Govt. schemes: - NREGA, SHGs, etc. -65 % of clients, but, only 51.7% of non-clients participate -opened various rural networks to generate product awareness; better product understanding Demographic characteristics: - Char. Demo. Correlation with take-up Joint family Positive SC/ST category Negative Household size Negative SMS Based Forecasts: Client Experience Salience of weather forecasts: -20% clients: weather forecast most important factor for good yield 0% Helps in managing risk of crop failure 15% -why weather forecasts? 48% 37% Helps in increasing yields of traditionally grown crops Helps in planting different variety of crops Other -top 3 desired qualities: regularity and availability in regional languages and trust on the source Experience with SMS forecasts: -only 6.16% clients were able to recognize weather related SMS -3.8% felt forecasts were regular and trustworthy -3.3% understood advice given -5.6% made recommendations Other sources of weather information Most used means for weather forecast TV forecasts Discussions with neighbours Respondents (%) 43.36 41 Radio forecasts 7.82 Traditional knowledge 6.16 Other means 0.95 SMS updates 0.71 Total 100 Impact of WII on clients Details on claims paid: -41 farmers had claimed Expected Avg. Crop premium compensation paid (Rs.) received (Rs.) Average Percentage compensation with received (Rs.) 30 % loss > Paddy Aman 482.1 226.7 1166.63 84.83 Paddy Boro 635.04 18.48 221.87 19.43 Impact on clients who claimed: -87.8% felt that the settlement was made on time; positive impact How was claim used (top three responses in order) How has claim helped (top three responses in order) Buy agricultural inputs Better off due to better production decisions in next season Buy agricultural equipment Better off due to presence of financial assistance in event of loss Repay loan Better off due to possibility of riskier decisions Impact on all clients: Contd. Impact of insurance on clients Percentage of clients Insurance helped in changing sowing decisions 23.7 Change in number of crops harvested 8.07 Helped in reducing farming costs 28.4 Good mechanism for coping with crop failure 83.9 Trusts the insurer 95.3 Faced difficulty in paying premiums 32.2 Weather insurance is very useful 94.3 -84% of clients feel insurance is a good mechanism to manage crop failure -94% feel insurance is useful -11.4% opened bank A/C after insurance purchase -95% trustWRMS as an insurance intermediary; positive impact Contd. Renewals: -Low renewal interest -32% found it difficult to pay premiums Reasons for non-renewal Percentage of clients Claims did not come through 80.57 Bad past experience 18.96 No trust/High basis Risk 0.47 Total 100 - majority of clients simply have not understood the product Reason for claim less than loss suffered Percentage of clients Related to work/location of weather station & basis Risk 8.06 All losses are not covered 0.94 Agents don't work 10.5 Don’t know 80.5 Total 100 Recommendations Client education and communication of product features Rural Networks used to spread product awareness Improved target group selection Reconsider mode, manner and delivery of VAS Adoption of flexible payment structure Thank You!!