Estimation of Willingness to Pay for Healthy Pistachio Consumption in Kerman

To respond to the more food need for the growing population of the planet, today, agricultural production using new technologies which have increased their products but this increase in production, has been faced with environmental problems such as pollution of soil and water resources, the development of new plant pests and diseases, malnutrition and disease due to decreased food quality. Nowadays, sustainable and organic food production seems vital more than ever facing with environmental, social, economic impacts of modern agriculture. Organizations and special policies have been formed for food production in developed countries. Organic production is associated with high price risk in Iran. So, in this study, the willingness to pay for healthy pistachio in Kerman is estimated to determine its true value using CVM to guide producers in order to evaluate the economic production.
In economic theory, changes in consumer welfare are estimated with surplus estimation and compensatory changes that reflect the willingness to pay for goods. One of the common ways to measure the willingness to pay is contingent valuation method. One of the most important parts of the CVM is to determine the extraction way of goods and services value. Information extraction method is different in CV studies including A) making recommendations, B) payment card (PC), C) Open -closed format (OE) and D) dichotomous choice (DC). Dual two-dimensional method (DB) is widely used with a great popularity due to statistical performance in recent years. Of course, there are some criticisms on this method that in some cases, it was observed that the answers to the first and second questions are incompatible. So, another form of dual method was introduced in 2002 by Cooper et al which is called OOHB which has been used in this study. Normally, in contingent valuation models, a parametric method with certain assumptions about the distribution function and form is used to estimate the parameters E.g.Logit and probit models. Parametric methods with assumptions based on proper function and accurate distribution cause efficiency estimators comparing to nonparametric methods, however, if these assumptions are not correct, it causes biased and inconsistent estimators. There are weak theoretical recommendations to select its function form and distribution. So in this study, the semi-parametric estimator of free distribution (SNPDF) is used to avoid estimator's inconsistency. Due to the above, the model used in this study follows as: In which the vector V includes variables such as income, age, education, family size, gender, family history of cancer and price. is a transfer function which in values shows the function less than 2л.This is necessary to prevent the rotation of the model. Therefore, the results not only show the variables of income, age, education, household size, gender, family history of cancer and the price but also show non-linear relationships between variables. Required data was collected through questionnaires and face to face interview with 87 people in 2014 and statistical analysis was done by GAUSS 9 software.
The results showed that 87 responders, 68 (78.16 percent) tended to pay more to buy healthy pistachio and 19 (21.84 percent) did not tend to pay more. 65.4 percent of people who did not tend to pay the proposed fee were women and 34.6 were men. The lowest and the highest bid in the questionnaire format was 380,000 and 550,000 Rials, respectively. The results of SNPDF using a questionnaire format of OOHB is given in Table 1.
Table 1. SNPDF model results to estimate the tendency to pay for healthy pistachio consumption
t-statistic Coefficient Variable
-1.195 -14.84 C
2.397 2.19 e-5 Age (A)
0.5549 1.461 Sex (S)
-2.928 -0.001918 Family size (N)
3.036 0.02268 Family history of cancer (C)
3.475 0.4901 Education level (E)
1.843 0.002289 Income (I)
-2.294 -0.001012 Bid price (B)
-1.967 -0.4419 uA
0.935 0.08898 vA
-0.829 -0.5585 uS
-2.748 -0.243 vS
-1.537 -1.967 uN
0.2351 0.935 vN
1.405 0.829 uC
0.2546 2.748 vC
0.8776 1.893 uE
0.157 3.485 vE
1.893 0.3157 uI
3.485 0.9575 vI
-1.825 -0.451 uB
-0.7364 -0.2914 vB
-709.3 Log-L
0.4147 McFadden’s
0.5256 Madalla’s
0.3889 Cragg&Uhler’s
486000 Average willingness to pay (Rials)
As you can see, the variables of education level, age, income, and family history of cancer, family size, and the proposal level are statistically significant. According to the results, the sign of bid price variable is negative and expected and represents that with the increase or decrease of bid price, the possibility of the bid acceptance decreases or increases. The age variable has a positive sign and shows that with the increase or decrease of age, the willingness to pay increases or decreases. The income variable has a positive and expected sign which means that with the increase or decrease of income, the possibility of the bid acceptance decreases or increases. The education level variable is positive and expected because by increasing the level of education, the importance of health care for people will be clearer. The variable coefficient of family size is estimated to be negative which shows that with the increase or decrease of family members, the possibility of willingness to pay decreases or increases. The variable coefficient of family history of cancer is positive and significant which shows that the higher family history of the cancer, the more willingness to pay. The story doesnt end here with semi- parametric method but the significance of some of uj and vj coefficients suggests that SNDP model could explain nonlinear relationships ignored in the parametric model.
As the results show, the average willingness to pay for a kilogram of healthy pistachio in Kerman is estimated 486,000 Rials while the price for a kilogram of non-healthy pistachio is 360,000 Rials on average. The tendency of people towards healthy product consumption shows that people care about their own health.