EMPIRICAL MODELS OF ECONOMIC CHOICE PROCESSES
MetadataShow full item record
The first chapter provides the first empirical attempt in using the new maximum likelihood-minimum power divergence (ML-MPD) binary response estimator. This nonparametric maximum likelihood estimator is used to model the underlying behavioral decision process leading to a person's willingness to pay (WTP) for recreation site attributes. The probit model and the Kriström/Ayer's estimators are also implemented. Based on the decision context and demographics of decision makers visiting the recreation sites, the ML-MPD approach suggests a more defensible representation of the underlying data-generating process and economic decision-making behavior. In the second chapter, a two-stage sequential experiment was conducted in a retail grocery setting to elicit WTP for four food products. In the first stage (round), participants bid on one of the four products. In the second stage, participants bid simultaneously for the other three products. The WTP for the food items was elicited using the Becker-DeGroot-Marschak (BDM) experimental auction procedure. I examine factors that may affect participants' bidding behavior across the two rounds. One factor is the uncertainty associated with the binding product in the second round, and the other one is a potential compensation effect on participants' bidding behavior across the two rounds. Results suggest that bids are sensitive to the context of bidding and to participants' preferences. Compensation has little impact on individual's bidding decision. However, there is some evidence that the uncertainty about which product will be binding in the second round, or the round order, can have an effect on participants' bidding decisions.The third chapter provides a first attempt at examining household purchase dynamics for dietary fiber, using a dynamic Tobit model that accounts for censoring across households and time as well as temporal correlations between current and previous purchases by adopting a stationary Gaussian first-order autoregressive choice process. Results indicate that household purchase decisions are characterized by significant unobserved heterogeneity, statistically significant positive serial correlation, and negative and significant state dependence, implying that lagged purchases have a strong effect on current household decisions so that households purchasing previously would buy less fiber in the current period.