| he wine industry has grown continuously over the past few decades in the United States and wine consumption has also increased during the same period. A survey study was conducted in order to identify the consumer base of wine market and winery tourism. A total of 701 consumers in North Carolina, New York, and Virginia were surveyed on their wine consumption patterns, visits to wineries, as well as demographics. The survey was conducted through telephone interviews from February through March in 2013. The first essay describes the results of the survey.;In the second essay, an empirical study is conducted to estimate the demand for wine. Using the survey data described in the first essay, two different empirical frameworks are applied. One is a continuous data model represented by the Tobit model, and the other is a count data model represented by the Poisson and the negative binomial models. The random effects specification is applied to each model in order to control for unobserved individual heterogeneity. As price data are necessary for every observation to estimate the demand models, unit price is imputed for the missing prices of zero purchase observations. Heckman's sample selection model and the ordinary least squares model are considered to estimate the unit price. Heckman's model is considered because the observations used to estimate the unit price is solely of those who did purchase wine and there could be sample selection bias. Since the results of Heckman's model indicate there is no sample selection problem, it is justified to use OLS.;Overall, the continuous and count data models show similar results. Along with wine price, factors that have a significant impact on wine demand are found to be household income and preference for wine taste. Although the estimates are somewhat different across models, individual characteristics such as level of education, race, and preference for certain color of wine have less impact on wine consumption. The price elasticity estimates vary between the continuous data model and the count data models. The price elasticity of wine demand is -1.4 in the random effects Tobit model, while it is -0.27 in the random effects Poisson model and -0.24 in the random effects negative binomial model.;In the third essay, a probit model is adopted to examine what affects an individual's decision whether or not to visit a winery in their state. The economic value wineries provide their visitors is derived from the winery trip demand that is estimated by the truncated negative binomial model. The results of the probit model show that what affect an individual's decision to visit a winery are household income, race, number of years lived in the state, and wine consumption. And the results of the truncated negative binomial model show that travel cost, household income, and age are the factors that significantly affect the number of visits to wineries. The consumer surplus estimates of visiting wineries are... |
