Friday, 6 January 2006 - 8:30 AM

Producer Preferences for Genetically Engineered Cotton in the Mississippi Delta

Swagata ″Ban″ Banerjee, Steven W. Martin, and Darren Hudson. Mississippi State University, P.O. Box 197, Stoneville, MS 38776

A mail survey of cotton producers in Mississippi was conducted in February and March of 2005 to determine the relative importance placed by producers on the attributes of cottonseed price, seed type (variety), lint yield, and fiber quality. Using these four attributes, a total of 26 scenarios were created based on statistical efficiency. They were divided into two blocks of 13 scenarios in each block, each scenario containing two alternative packages (A and B) of specified levels of each attribute, and the option of choosing none of the two packages (choice C).

A questionnaire consisting of 13 scenarios and some demographical questions was sent to each of the 600 cotton producers (300 receiving each block) selected randomly using a simple random number generating technique by the computer. These 600 producers were selected from a possible list of 1,319 cotton producers in Mississippi sent to us by the Northwest District County Extension offices.

Out of the 600 questionnaires mailed, three were returned undeliverable, and a total of 394 questionnaires were not returned. Of the 203 questionnaires that were returned, 86 could not be used because 83 of them indicated the respondents were either not cotton farmers or had retired, and three of them were returned blank. Therefore, a total of 117 (= 203-86) cotton producers' responses were usable for this study as they provided data. Assuming the non-respondents to the survey were active cotton producers, the, the overall response rate was approximately 34% (= 203/597), and the usable response rate was approximately 22.9% (= 117/511).

Assuming a cotton producer's decision to purchase cottonseed depends on the attributes of seed price, seed type, lint yield, fiber quality, and other considerations, the survey asked the respondents to evaluate on a scale of 1 to 5 (1 being very important to 5 being very unimportant) how each of these attributes would influence their decision when purchasing cottonseed. Based on the responses received, “other considerations” received the highest scores, viz., mean, median and mode of 2.53, 2.50 and 3.00, respectively. The mean and median for seed price scored a close second, viz., 2.28 and 2.00, respectively. The means of fiber quality, seed type and lint yield were in third, fourth and fifth places, respectively, with scores of 1.82, 1.68 and 1.60 on the chosen 5-point scale. Mode for all attributes except “other considerations” was 1.00.

Regression results from a conditional logit model indicate respondents preferred "none" to either of their available package choices (as given by the negative sign on the constants). Coefficient on price is negative, as expected. Further, marginal willingness-to-pay approach in conditional logit shows there is a positive willingness to pay for technology relative to conventional cotton, and it increases with the level of technology. Yield and fiber quality each has a positive willingness to pay, and the higher the quality, the higher is the willingness to pay.


[ Recorded presentation ] Recorded presentation

See more of Cotton Production Economics Session
See more of Cotton Economics and Marketing Conference

See more of The Beltwide Cotton Conferences, January 3-6 2006