17 Demand for drought insurance in Ethiopia

Transcription

1 128 The challenges of index-based insurance for food security in developing countries 17 Demand for drought insurance in Ethiopia Million Tadesse (1) (2), Frode Alfnes (1), Stein T. Holden (1), Olaf Erenstein (2) Corresponding author (M. Tadesse): 1 Norwegian University of Life Sciences (UMB), School of Economics and Business, PO Box 5003, Ås Norway. 2 International Maize and Wheat Improvement Center (CIMMYT), C/o ILRI PO Box 5689, Addis Ababa, Ethiopia. KEYWORDS: Ethiopia, weather index insurance, choice experiment, insurance for work, econometrics SUMMARY Although food-for-work or cash-for-work programmes are popular in many developing countries, the idea of work for insurance is new. Using a choice experiment approach in the Southern Nations, Nationalities, and Peoples Region (SNNPR) and Oromia Region of Ethiopia, we found that smallholders are interested in these types of insurance products and insurance premium payment vehicles, and that most of them are willing to participate in the insurance-for-work programme at lower daily wage rates than those commonly found in food-for-work or cashfor-work safety net programmes. However, most are financially restricted to paying premiums in cash unless subsidised. 1. Introduction One in every six countries in the developing world faces food shortages because of droughts (Liu et al., 2008). In recent years, the introduction of weather index insurance schemes has received increasing interest due to new ways of controlling moral hazard and growing concerns about increasing climate risks due to global warming. Despite the various benefits of weather index insurance in agricultural risk transfer and mitigation, results from various pilot studies indicate that the uptake of weather index insurance products by smallholder farmers is below expectation (Dercon, 2002; Suarez and Linnerooth-Bayer, 2010; Giné et al., 2008). High premiums have been one of the key factors negatively influencing the demand for weather index insurance in low-income countries (Carter, 2009). Although price is an important factor affecting demand, it is not the only factor (Cole, 2012). Product design, basis risk, trust and perception of insurance principles by the poor are equally important (Cole, 2012; Carter, 2009).

2 129 The existence of social networks, government subsidised post-disaster management strategies such as food-for-work or cash- for-work programmes may affect the demand for weather index insurance. This is because the above strategies may be cheaper for the poor as long as the strategies are accessible. Unfortunately, the use of informal risk management strategies and that of government post-disaster management are often costly. This is because such strategies often arrive after households have sold their assets and, therefore, recovery would be difficult. In this regard, provision of low cost weather index insurance (based on records of local rainfall station) is expected to enhance product uptake despite some of its inherent problems such as basis risk. In this study, we allow households the possibility of paying their insurance premium in the form of labour, a combination of labour and cash, and cash only to determine households willingness to pay for drought insurance. Payment with labour has previously been possible in food-for-work programmes, while work for insurance is a new idea originally suggested in Tadesse and Victor (2008). 2. Experimental Design Households were asked to choose between three insurance schemes and a none-of-these option. The insurance schemes were described by three attributes: work, cash, and payout. To secure identification of the attributes we used the design software Ngene (3) to create fractional factorial design with three factors: work, cash, and payout rate. We analysed the choice data using a mixed logit model where we allowed the parameters of all three attributes in the choice experiment to be random parameters with a normal distribution. To capture how previous events related to food shortages, access to safety nets and risk attitude affects the willingness to buy insurance, we added three dummy variables as interaction terms to the two payment variables: work and cash. The work attribute was the number of days per year the participants had to work to buy the insurance in a work-for- insurance programme. The cash attribute was the amount per year the participants had to pay to buy the insurance. The payout was the amount of money the participants would receive if there is as little rain (based on a nearby rainfall station record) as is typical every fourth year. The expected yearly payout would, therefore, be 25 % of the payout. Each attribute has five levels in the choice experiment. 3

3 130 The challenges of index-based insurance for food security in developing countries 3. Data and results This study was conducted in two districts in the Southern Nations, Nationalities, and Peoples Region (SNNPR) (Damot Sore and Wondo Genet) and two districts in the Oromia Region (Sashemene and Arsi Negele) of Ethiopia (Figure 1). Figure 1: four districts in SNNP (Southern nation s nationalities and peoples: Damot Sore and Wondo Genet) and Oromia (Sashemene and Arsi Negele) These districts represent different agroecologies, production potential and access to markets. A total of 576 randomly sampled households were used for our choice experiments and answered six or eight choices each. This makes our total number of choice observations In addition to the information on drought insurance, other information including exposure to shocks, access to Productive Safety Net Programmes (PSNP), potential risks and coping mechanisms was collected. A typical farmer in the most densely populated areas (Damot Sore District) has an average farm size of 0.5 ha to support a family of five persons. Only a handful of farmers in Wondo Genet District use irrigation, otherwise the majority of our farmers use rain for crop and livestock production. Hence, the effect of drought is very significant in these farming systems. Participants familiar with micro-finance and iddir (a funeral society) quickly recognise the inherent merits of insurance, while participants who had no experience of these institutions struggled at the outset of the focus group discussions to grasp the concept of formal risk transfer (about 37 % of the sample farmers had not previous heard of insurance). When asked whether they would demand their premium be refunded in periods of no loss based on a weather station record not necessarily on their farm, the majority stated that they would not demand a refund (63 %) but 5 % stated that they would demand their premium back. About 87 % of the sample households considered the possibility of contributing their labour in tree planting or other development activities to receive insurance cover as a win-win approach.

4 131 This is because tree planting in itself helps to conserve soils that could help increase crop productivity and allowing labour as payment for insurance premiums also helps transfer risk to the third party (insurers). About 75 % of households experienced food shortages within the last four years. Both late arrival and early end of the rain are commonly mentioned as the major problem compared to a totally dry spell. Some of the findings from our mixed logit model indicated that respondents, on average, were subject to the same disutility from working one day as from paying ETB 2.43 (4), and that they are willing to pay ETB 0.81 for an expected yearly payout of ETB 1. We note two things: firstly, the implicit wage rates are much lower than the current wage rates from food-for-work programmes in Ethiopia; secondly, the average participants need a subsidy to pay cash for insurance since their willingness to pay is less than the expected cost of the insurance for an insurance company. 4. Conclusions This paper uses a choice experiment to elicit smallholders willingness to pay for weather index insurance in the highlands of Ethiopia. Although food-for-work or cash-for-work programmes are popular in many developing countries, the idea of work for insurance is new. Using a choice experiment approach in the SNNP and Oromia Regions of Ethiopia, we find that smallholders are interested in these types of insurance products and insurance premium payment vehicles, and that most of them are willing to participate in the insurancefor-work programme at lower daily wage rates than those commonly found in food-for-work or cash-for-work safety net programmes. However, most are financially restricted and are not willing to buy insurance with cash unless the insurance is subsidised. Households exposed to food shortages in previous seasons are more willing to work for insurance than those that have not experienced food shortages; however, they are not willing to pay more in cash. Hence, interlinking traditional risk management practices (such as PSNP) with insurance will provide a win-win outcome. This is because even in those periods with no payout, farmers may still benefit from the trees planted/soils conserved. Insurance providers are also benefiting from the premium collected in this non-drought period. Furthermore, as smallholders are interested in working for insurance certificates at lower wage rates than the current PSNP rate, this means that the difference in wage rates could be used for insurance subsidies with no additional cost to donors or the government to subsidise the premiums. 4 ETB (birr) is the Ethiopian currency: UDS 1 = ETB 17.96, 2 October 2012.

5 132 The challenges of index-based insurance for food security in developing countries 5. References Carter, M. (2009), Intelligent Design of Index Insurance for Smallholder Farmers and Pastoralists, Innovations in Insuring the Poor, Focus 17, Brief 6, International Food Policy Research Institute. Cole, S., Giné, X., Tobacman, J., Topalova, P., Townsend, R., Vickery, J. (2011), Barriers to Household Risk Management: Evidence from India, Harvard Business School Finance Working Paper , Harvard University, Cambridge, MA. Dercon, S. (2002), Income risk, coping strategies and safety nets, The World Bank Research Observer, 17 (2): Giné, X., Townsend, R., Vickery, J. (2008), Patterns of Rainfall Insurance Participation in Rural India, The World Bank Economic Review, 22 (3): Liu, J., Fritz, S., van Wesenbeeck, C. F. A., Fuchs, M., You, L., Obersteiner, M., Yang, H. (2008), A spatially explicit assessment of current and future hotspots of hunger in Sub-Saharn Africa in the context of global change, Global Planetary Change, 64(3 4): Suarez, P., Linnerooth-Bayer, J. (2010), Micro-insurance for local adaptation, Wiley Interdisciplinary Reviews: Climate Change, Focus Article, 1 (March/April), Tadesse, M., Victor, M. (2008), Estimating Demand for Microinsurance in Ethiopia, Oxfam America first report, commissioned by ILO/UNCDF, Geneva, Switzerland. 6. Acknowledgement The authors gratefully acknowledge financial support from the Research Council of Norway (Project No ) and the CIMMYT Socioeconomics programme for supporting the research in Ethiopia. The research was carried out while the first author was a postdoctoral research fellow at UMB and visiting scientist at CIMMYT.