Consideration of zero values in willingness to pay: A case study of rice farmers in the Korhogo area

Abstract

The econometric treatment of zero values in the willingness to pay studies has sparked debate among economists. This is an important question since the scope of recommendations depends largely on the results obtained. The objective of this article is precisely to make a comparison between the Tobit and Heckman models to see which one allows to better take into account the zero values. To achieve this, we conducted a contingent valuation study by studying the willingness to pay of Korhogo rice farmers for rainfall deficit index insurance. Based on a questionnaire, we collected data from 269 rice farmers in the Korhogo area. The results indicate that the average willingness to pay is around 5,500 CFA francs. Moreover, if rice farmers have a good knowledge of insurance (55.39%), they seem not to know agricultural insurance (1.41%). In addition, the answer to our comparison problem of Tobit and Heckman shows that Tobit model seems more appropriate because of the low value of Akaike's criterion (3632,011) compared to Heckman's (4393,42). The study suggests that the index-based agricultural insurance market must be developed in Côte d'Ivoire to help preserve farmers' incomes. To do this, it is essential to popularize this product among farmers and make them aware of the importance of insuring their farms. Finally, the choice of Tobit seems more appropriate in our case since it better accounts for the treatment of null values. It is therefore important that other studies be conducted with complementary tests in order to properly treat zero values in willingness-to-pay studies.

Keywords : Contingent Valuation, Willingness to Pay, Index Insurance 

JEL Classification : C24, Q12

Paper Type : Empirical Research