Predicting Crop Insurance Share of Preference in Tanzania Using Randomized First Choice Model (RFC)

Abstract

Godwin Stanslaus Kalokola*, Arbogast Moshi and Joel Johnson Mmasa

There are four methods of market simulation models, such as share of preference, Randomized First Choice (RFC), first choice, and total utility, which are popular in different countries for predicting crop insurance share of product. In Tanzania literature is scant in all methods. This paper only focuses on the Randomized First Choice (RFC) as suggested by Orme (1998) and later refined by Huber, Orme, and Miller (1999) for predicting crop insurance share of preference. The Choice Based Conjoint (CBC) study was used to investigate preferred crop insurance packages based on three developed hypothetical crop insurance products, i.e., Najivunia, Fahari Kilimo, and Siteterki, with four attributes and three varying levels. An eight-choice task with a combination of random and fixed holdouts was shown to 360 maize smallholder farmers in Kongwa District to elicit preferred or highest crop insurance with high utilities. Hierarchical Bayes analyses were applied to compute utilities, and a market choice simulator was used to predict choices of the highest product for different combinations of crop insurance defined using the attributes and levels.

Results: The RFC shows smallholder farmers have high utilities for the Najivunia package of crop weather index insurance with 46 average utilities and a 73% share of preference. The package offers farmers the option to buy the policy using a mobile phone; the suitable time to pay the premium is when farmers have harvested the crop, and to pay the premium in installments rather than other options, with the incentive of input credit or a loan bundled together as an incentive, while the package offers an affordable premium of TSH 10,000. The result shows an introduction of a new product with modification of attributes of high interest to farmers with both product and price sensitivity will not substantially reduce the share of preference to 30%, showing stable preference of the package. Digitization and digitalization are among the key drivers for these packages to reach large segments of rural farmers who have experienced adverse weather events, where crop insurance is an ideal tool for protection, but big agriculture data to enable pricing and product development of an index product correlates with farmers’ losses at various locations.

HTML PDF