Modeling Revenue and Visitation Patterns of Agritourism Operations in Oklahoma

Abstract

With agritourism increasingly promoted as a way to generate income in rural areas, information is needed on the potential performance of agritourism operations. Currently, existing research provides little information about the business characteristics that can affect agritourism revenue. This manuscript presents research on the business characteristics associated with differences in agritourism revenue of Oklahoma businesses. This research also identifies whether offering wedding services financially benefit agritourism businesses. Ordinary least squares (OLS) and interval regression techniques are used to estimate the revenue model. In the latter case, tests are performed to gauge the effects of different interval sizes on model estimates. There is also a lack of information relating the use of different marketing methods and agritourism visitation. This manuscript presents research examining the effect that different marketing methods have on the number of visitors a business receives. A visitation model is developed that measures the significance of seven distinct marketing methods. Three different specifications of the visitation model are considered: a linear model, a log-linear model and an exponential model estimated using a quasi-maximum likelihood (QML) estimator. The exponential method is based off research performed by Santos Silva and Tenreyro (2006) that shows log models can produce inconsistent parameter estimates when heteroscedasticity exists.