Located in a low latitude belt of the northern hemisphere, Taiwan is right in the path of Western Pacific typhoons. The climate here is complex and diverse. In summer and fall the island is subjected to frequent typhoons and heavy rains, while in winter there are low temperatures and damage from cold. The combination of the geographic location and the topography leads to frequent natural disasters. In recent years, as the global climate has changed, weather events have become more extreme, and the severity and frequency of natural disasters have both increased continually. In addition, unusual temperatures and rainfall have created higher risk for agricultural production. It is no longer possible to protect the incomes and assets of farmers purely by relying on government-funded disaster relief.

　　Agricultural insurance is based on the principle, underling all insurance, of bringing together large numbers of farmers to share the risks from natural disasters, and thereby cover losses caused by their impact. Taking into account the methods and scale of operations in the insurance sector, and the level of difficulty involved in assessing losses and arranging for compensation, the COA decided to begin by offering only livestock insurance. In order to uphold the rights and interests of farmers, the COA also, starting in 2007, began commissioning scholarly bodies to undertake studies to research feasible methods to undertake "natural-disaster insurance for agriculture" (hereafter NDIA). Moreover, in 2014, we undertook an analysis of the feasibility of insuring specified crops of high economic value, drafted possible insurance contracts, and calculated premiums; these are all essential to evaluating the possible implementation of NDIA on a trial basis.

　　Because Taiwan has a wide variety of crops, the average cultivated area per farmer is small, and crop cultivation areas and seasons are both concentrated, risk diversification is difficult. The structure of agriculture does not conform to the basic insurance principle of the law of large numbers. Moreover, insured crops should conform to the insurance principle of "low incidence, high loss," and the number of insurees should be sufficiently high to diversify risk, because only then will commercial insurance companies be willing to participate and make NDIA effective.

　　In order to undertake trial implementation of NDIA, in 2014 the COA, for the early stages of policy planning, evaluated which crops in Taiwan are most feasible for participation in the first phase. Priority consideration was given to those crops which have high economic value and which have attained a certain scale of production. After inviting the private sector and research institutions for discussions, and gaining a deep understanding of the situations of various crops, 10 crops were selected as candidates for suggested trial implementation of NDIA in the first stage, including: top-grafting pears, mangoes, sweet persimmons, Kyoho grapes, papaya, citrus tankan (tonkan), wendan pomelo, custard apples (sugar apples), wax apples, and wet paddy rice. We also asked the Non-Life Insurance Association of the ROC to notify member companies about trial NDIA to start in 2015, and to encourage property insurance companies to actively participate.

　　The COA then gathered relevant information— including the cultivated area, geographic distribution, annual production value, number of farm households involved in cultivation, and disaster-relief payments in recent years—for each of these crops. We asked the government agencies with jurisdiction over the insurance industry to seek out opinions on issues connected to NDIA of these crops. Also, based on data on production costs gathered in surveys over the years, we asked for the opinions of relevant research and experimental units, and made a detailed chart of costs required for each cultivation season. It is hoped that providing comprehensive information on these crops to private-sector insurance companies will increase the willingness of such firms to offer these insurance policies.

　　The COA plans to begin trial implementation of NDIA on a commercial insurance basis in 2015. We hope to draw on the efficiency of private insurance firms to rapidly make NDIA available. The experience gained in the trial implementation period can serve as a reference in the future for system design, drafting of laws and regulations, insurance premiums, underwriting methods, and risk diversification mechanisms.