Font Size: a A A

Tree fruit market analysis with limited dependent variable models

Posted on:2014-12-26Degree:Ph.DType:Dissertation
University:Washington State UniversityCandidate:Li, HuixinFull Text:PDF
GTID:1453390008460919Subject:Economics
Abstract/Summary:PDF Full Text Request
This dissertation investigates the tree fruit market using limited dependent variable models. The crops of interest for this dissertation include apple, peach, strawberry, sweet cherry and tart cherry. The dissertation is composed of three related studies. A consumer demand for apple varieties is analyzed based on household purchase dataset. Market intermediaries' willingness to pay and priority setting of genetic fruit traits are estimated by discrete choice models using survey data.;The first study applies Almost Ideal Demand System to investigate consumer demand for five fresh apple categories. Using Nielsen Homescan cross-sectional data from 2005, an Expectation Maximization approach is adopted to correct for censoring problem in multiple equations. Elasticities are obtained and used to explain the relationship of five varieties. The results indicate that demand for organic apples is more sensitive to price change than conventional apples. Only substitute relationship exists between different apple varieties.;In the second study, we designed choice experiment surveys to investigate market intermediaries' willingness to pay for increments in targeted genetic traits of rosaceous fruit. The choice data are analyzed using discrete choice models including conditional logit model, heteroscedastic extreme value (HEV) model, mixed logit model and multinomial probit model. The results suggest that market intermediaries' highest WTP for genetic traits differs for each rosaceous fruit crop.;The third study conducted surveys to elicit market intermediaries' rating of importance for various fruit quality traits. In the survey, market intermediaries are asked to rate the importance of listed traits using a 7-point Likert scale. Since the ratings of importance for genetic traits have a natural ordering, a multivariate ordinal probit model is appropriate to analyze the data. The results show that the most important trait is storage life for apples, pit removal for cherries, flavor for peaches and flavor for strawberries.
Keywords/Search Tags:Market, Fruit, Model, Using
PDF Full Text Request
Related items