Economics of Weather Index-Based Insurance: Analysis of Smallholder Farmers' Preferences and the Impact of Insurance on Productivity in Kenya
von Kenneth Waluse Sibiko
Datum der mündl. Prüfung:2016-11-10
Betreuer:Prof. Dr. Matin Qaim
Gutachter:Prof. Dr. Xiaohua Yu
Gutachter:Prof. Dr. Oliver Mußhoff
EnglischClimatic risks are a major challenge for rural communities, especially those whose livelihoods directly depend on agriculture. Weather extremes affect crop productivity not only through the direct relationship between weather variables and yields, but also through the negative effects of such shocks on farmers’ investment behavior. In particular, smallholder farmers are vulnerable partly because they are predominantly located in the tropics where exposure to climate variability is high, but also because they lack the resources and the capacity to properly cope with risks. Agricultural insurance is one possible solution, but for long formal crop insurance remained unavailable in most developing countries. Institutional constraints, including high transaction costs, the spatially correlated nature of agricultural risks, and classical issues of information asymmetry, discouraged insurers from offering agricultural insurance contracts. Recently there have been efforts to change this situation through the development and provision of weather index insurance (WII) contracts. A WII contract is advantageous because the insurer relies on an objectively measured weather variable that is correlated with farm losses but which cannot be manipulated through farmers’ actions. Usually indexes used in WII are measured at the aggregate level, minimizing issues of moral hazard and adverse selection into insurance programs. Moreover, the insurer is saved from having to assess losses on each insured farm, an aspect that cuts down administrative costs. In turn, if the WII contracts are well designed, insured farmers can benefit because of relatively affordable insurance premiums and faster claim payments in case of a devastating weather event. Despite these potential benefits, demand for the piloted WII products has generally been low. Farmers’ preferences have rarely been analyzed; hence it may be that existing WII contracts are not sufficiently tailored to the needs of smallholder producers. Understanding the main obstacles from the farmers’ perspective can help in improving the design and implementation of WII schemes. Furthermore, WII initiatives are usually intended to contribute towards poverty reduction in the long run. However, very little empirical evidence is available in the literature on the economic and social impacts of existing WII programs. To address the mentioned knowledge gaps, in this study two research objectives are pursued, each one representing one of the main chapters of this dissertation. The first objective relates to farmers’ preferences for WII. We combine data from a survey and a discrete choice experiment (DCE) conducted in Kenya to analyze farmers’ experience with an existing WII program and preferences towards specific improvements in WII contract design attributes. Specifically, the DCE included price and non-price attributes related to proximity to weather stations, rainfall thresholds, insurer transparency, and group rather than individual contracts. Kenya is an interesting setting for the study because farmers there already had some experience with WII. In particular the so-called Kilimo Salama program has been in existence since 2009, providing index-based crop insurance products in different parts of the country. For the analysis we use a mixed logit model that accounts for preference heterogeneity given that farmers may not always portray similar preferences. Results indicate that the existing WII contracts are probably too expensive from farmers’ point of view. This could be one reason why insurance demand remains low. Besides affordability, many smallholder farmers do not fully understand how the insurance functions. Transparent provision of information on realized rainfall measurements and threshold levels would improve farmers’ willingness to pay for insurance. Mechanisms that improve the effectiveness of WII contracts, such as increasing the number of weather stations, are also valued by farmers, but not to the same extent as frequent communication. In addition, offering WII contracts to groups – rather than individual farmers – could be a promising avenue for achieving more widespread uptake of WII among smallholder farmers. Group contracts could reduce transaction costs. They could also offer important platforms through which farmers can learn and better understand the complexities of index insurance contracts. The second research objective pursued in this dissertation relates to the effects of WII adoption on input use and crop productivity. Weather risk has been shown to affect the usage of productive inputs including fertilizers and modern seed technologies. Risk-averse farmers are usually reluctant to use external inputs, when rainfall and temperature uncertainties are too high. However, because of this tendency of avoiding risks, they also forego potential yield gains that they could realize if they utilized more external inputs. Using randomized field experiments, a few recent studies have analyzed whether the provision of WII can improve investments in riskier but more productive technologies. Unlike these previous studies, we use observational data from a survey of smallholder farmers, analyzing the effects of actual participation in the Kilimo Salama WII Program in Kenya. For the analysis we utilize a treatment regression model with instrumental variables. Empirical results reveal a significant increase in the use of fertilizer and improved seeds as a result of adopting WII. Further analysis also indicates a significant improvement in maize productivity, which mainly comes from the increase in fertilizer application. The results emphasize that WII is indeed one mechanism for promoting productivity growth, however, this potential is not yet fully realized, because only a small proportion of farmers has so far participated in the WII Program. Overall, the study concludes by stressing the need to educate farmers about novel index insurance products that they may not be familiar with. Although some training is currently being provided as part of Kilimo Salama, this has to be intensified to inform farmers on potential benefits and limitations of the WII products. For this it may be useful to harmonize insurance trainings with public agricultural extension, so as to improve the effectiveness of the training interventions. Strengthening producer groups and encouraging farmer-to-farmer knowledge exchange within group networks are also other ways to improve the rate of insurance uptake. WII is definitely not a silver bullet for productivity growth in the small farm sector; however, providing effective WII contracts jointly with other support services may contribute positively to the well-being of risk-prone farming communities.
Keywords: Climate risk; smallholder farmers; fertilizer; maize productivity; crop insurance; discrete choice experiment; endogenous treatment regression; Africa