«Der Open-Access-Publikationsserver der ZBW – Leibniz-Informationszentrum Wirtschaft The Open Access Publication Server of the ZBW – Leibniz ...»
This study is part of a research project on the analysis of requirements for the development of an economically-sustainable crop insurance in a transition country using the case of Kazakhstan. Particularly, the objective of this study is a comparison of several insurance schemes with respect to their ability to serve as an acceptable instrument of farm income stabilization in transition. The assessment is based on both the literature on the issue and the preliminary results of a numerical analysis of farm and weather data. The study uses extensively the results and data from a farm survey conducted in the framework of the project (HEIDELBACH et al., 2004). The author thanks Olaf Heidelbach for his helpful assistance in preparing chapters 2 and 3 of this discussion paper. A special word of thanks goes to the project associates Bota Borina and Darina Ostrikova who were extensively involved in the data collection. The author is also grateful for advice provided with regard to drought index application by Alexej Ivannikov from Agrarian University in Astana, Ludmila Chuntonova from Kazhydromet (Kazakh Hydro-meteorological Agency) and Irina Yesserkepova from Kazakh Research Institute for Environmental Monitoring and Climate.
The paper is organized as follows: Section 2 gives a short overview and systematization of the most current and widespread insurance products. Section 3 presents a discussion of the comparative advantages of two well-established and two relatively new crop insurance schemes.
This discussion is followed by a quantitative assessment of the potential for introducing parametric (index) insurance in Kazakhstan. Conclusions are drawn in the final section.
2 SHORT OVERVIEW OF INSURANCE PRODUCTSCrop insurance is used in many countries and a variety of crop insurance products are offered worldwide2. Several relatively new insurance schemes are being investigated to respond to special needs and issues on pilot-basis. The diversity of insurance products makes it difficult to draw a clear distinction between them. Therefore, before starting an analysis of different insurance schemes, the most important insurance products will be presented and systemized to provide an understandable overview (Table 1).
Generally, one can distinguish between all-risk, multiple risk and particular risk insurance.
Two additional important groups of insurance schemes should be considered separately:
Parametric and catastrophic insurance. At the same time, two mechanisms of crop insurance could be distinguished. The first mechanism is based on the actual production history (APH) of the farm. APH provides the base for different calculations using the insured’s historical yield records. Another mechanism of insurance is the so-called parametric or index-based insurance, which uses weather or area-yield indexes for pricing insurance contracts. Thereby, insurance payoffs are subject to the occurrence of a special weather event, which can be described by a weather-based index (SKEES, 1999). In case of area-yield insurance, average area yield "triggers" an indemnity payment which is equal to the difference, if positive, between the annual area yield and some predetermined critical yield (MIRANDA, 1991).
3 QUALITATIVE COMPARISON OF INSURANCE PRODUCTSIn light of the complexity of challenges and many interdependencies between individual aspects of insurance market development, it is important to set up criteria which can help to compare individual insurance products. Though it is not easy to draw a clear division between
individual aspects, the following assessment features were considered in this study:
• Incentives for farmers to buy insurance;
• Incentives for private insurance to provide crop insurance;
• Possible effects on productivity and production patterns;
• Feasibility (applicability);
• Financial viability of insurance scheme.
In addition the assessment considers several issues which are especially important in the
transition context. The most important follow:
• Underdevelopment of financial markets;
• Possible presence of marginal production areas, and hence a higher exposure to systemic risks, which can seriously affect the development of financially-viable crop insurance;
• Large differences in farm productivity that can induce adverse selection;
• Information deficiency in view of complex farm restructuring and changes in production patterns;
• Underdeveloped market infrastructure, which lowers the profitability of farming;
• Low liquidity of farms, which can hinder their participation in crop insurance schemes;
• Many farmers had bad experiences with insurance during the Soviet era. This makes them cautious and less interested in insurance;
• Low attractiveness of involvement in agricuture on the side of private insurance, first of all due to high risk and transaction costs. However, not least due to low profitability of farming in general.
A short presentation of the particular advantages and disadvantages of the considered crop insurance schemes with respect to the selected criteria is provided in Appendix B. Several issues, however, will be more precisely examined in the following.
Past experience strongly suggests that not all risks are insurable. In agriculture in particular, many crop insurance programs fail to operate on an actuarially-sound basis. In theory, there are two attitudes towards the question of risk insurability. Among others, BERLINER (1982) underlines the requirement that it must be possible to make reliable estimates of the relevant probabilities from statistical observations. The implication is that a risk is insurable only if it can apply the law of large numbers. In the insurance sector, risk is classified as insurable as long as actuarially-sound premiums are charged. Actuarially-sound premiums have to accurately reflect the risks involved. However, actuarially-sound premiums can often be established only Raushan Bokusheva at a very high premium or cannot be achieved at all (MEUWISSEN et al., 1999). With respect to the realization of the law of large numbers, a serious difference may be constituted only regarding two general options: Compulsory and voluntary insurance. In principle, every insurance product considered in this analysis can be provided in one of both ways. As mentioned before, many socialist countries tried to realize insurability by introducing compulsory insurance. Farms had to pay for insurance without any decision option (even if they did not need one). Moreover, the premiums established by the state insurance companies were not correlated with the actual risks involved, as premium rates were distinguished only according to relatively large territorial units (ZADKOV, 1997; PYE, 2000). Such developments induced negative experiences with insurance in the cases of successful enterprises and the free-riding behavior of loss-makers. The process of privatization in Kazakh agriculture has had a significant impact on the importance of risk for agricultural producers. Nowadays, farmers inevitably have to adapt their production to natural production conditions (PETRICK, 2001). Thereby, they are looking for appropriate instruments of risk mitigation. As the results of a farm survey3 show, 64.4 percent of the respondents would like to be insured. However, only 43.8 percent of this number believes that crop insurance should be compulsory in Kazakhstan (HEIDELBACH, BOKUSHEVA and KUSSAYINOV, 2004). A compulsory insurance scheme usually undermines the farmer’s decision-making autonomy and hence affects activity of individual farmers. In such circumstances, farms are forced to employ risk-management instruments which may not provide the best solution to the farm's problems, or must even pay for services which they do not need. This makes compulsory insurance rather different from transition goals, since it violates free decision-making and, respectively, production factors allocation. Additionally, a compulsory insurance scheme is usually heavily regulative, which prevents insurance companies from setting actuarially fair premiums.
In addition, to realization of the law of large numbers, the literature specifies two further aspects that have an effect on insurability: Systemic risk and asymmetric information. In assessing the insurability of risks in agriculture, MIRANDA and GLAUBER (1997) identify both as basic conditions for risk insurability: First, the risks should be nearly stochastically independent across insured individuals; second, the insurer and the insured should have very nearly symmetric information regarding the probability distribution of the underlying risk.
Contrary to automobile or fire risks, which tend to be independent, the crop-yield risk exhibits a substantial degree of correlation across space (MIRANDA and GLAUBER, 1997). As stated before, crop losses in Kazakhstan are often driven by natural disasters, which simultaneously affect a large number of farms over a widespread area. Drought and extremely high temperatures are the main natural hazards that induce systemic yield losses of grain producers in most important production areas there. In light of the high specialization scale of Kazakh agriculture, where grain currently makes up 80 percent of gross agricultural output and covers 79 percent of the sown area in Kazakhstan (STATISTICAL YEARBOOK KAZAKHSTAN, 2003), the problem of systemic risk can be especially serious. The concentration of grain production in the northern regions in Kazakhstan with similar climatic conditions makes this issue even more severe (Appendix C demonstrates the correlation of regional grain yields). In this context, considering the capacity of an insurance scheme to treat systemic risk is of great importance in comparing alternative insurance products. As multi-peril yield and revenue insurance could not provide a solution for systemic risk, innovative insurance schemes have been considered in several countries. Currently, area-yield insurance and weather-based insurance are regarded as This farm survey was implemented in October-November 2003 and May-June 2004. 73 farmers and managers of agricultural enterprises were interviewed in the different parts of the country during this time (HEIDELBACH et al., 2004).
Crop insurance in transition: a qualitative and quantitative assessment of insurance products 13 the most appropriate alternatives to conventional insurance products. However, the high correlation among individual farm-level yields maY force insurers to charge a high risk premium which makes insurance unattractive (MAHUL, 2001). The problem in this context is that risk pooling is difficult to achieve between those who are exposed to the same type of systemic risk.
Hence, to manage the problem of systemic risk in agriculture, risk pooling must be extended to other economic sectors, for example, by introducing financial market products such as weather derivatives. At the same time, considering the case of a transition country requires much attention to be paid to the economic viability of agricultural production in individual regions. If long-term farm profitability is not achievable due to unfavorable weather and production conditions in a region, risk pooling would not be an appropriate mechanism of farm income stabilization, since it would imply an income redistribution from profitable to unprofitable farms and, respectively, from more productive to less productive sectors of the economy.
Asymmetric information manifests itself primarily in terms of adverse selection and moral hazard. Adverse selection in insurance markets is caused by the inability of the insurer to accurately rate the risk of loss of individuals who purchase insurance. Moral hazard is a result of hidden actions of the insured, which increase the risk of loss of the insurer. Theoretical and empirical studies (AKERLOF, 1970; ROTHSCHILD and STIGLITZ, 1976; MAKKI and SOMWARU,