Consumer Preferences and Willingness to Pay for Natural Beef: a Discrete Choice Experiment Approach

“Natural” is one of the most common words appearing on new food products. Despite the wide use of the term, the Food and Drugs Administration (FDA) has not provided a formal definition of natural. The absence of a clear definition has led to various citizen petitions, either to define the term or even to prohibit its use. The main argument for prohibiting natural labeling is that the term is potentially misleading. Furthermore, findings in the existing literature indicate that some consumers tend to confuse natural and organic labels. A misleading label could lead to distortion of the consumer’s budget allocation. Meanwhile, a non-misleading label would not affect consumer preferences and willingness to pay (WTP) for the product. Given this, in the second chapter of my thesis I examine if consumers that are aware of the United States Department of Agriculture (USDA) definition of natural and organic exhibit different WTP for natural beef in comparison to consumers that are either unaware or unfamiliar with the USDA definitions. Findings suggest that consumer knowledge regarding the definition of natural affects consumer WTP for beef. Furthermore I examine the complementation and substitution among different labels, providing useful implications to producers regarding the benefits of labeling their products. In the third chapter, I evaluate if the use of verbal or pictorial representation of the alternatives in a choice experiment affects the responses of the participants. Findings indicate that the structure of the utility functions are different among the pictures versus text representation of choice sets. This phenomenon is further illustrated by modeling for attribute non-attendance (ANA). When ANA is included in the model, the utility functions resulting from visual presentation, resemble with those coming from the text presentation. On the other hand, text representation of choice sets results in more statistically significant WTP estimates than the use of photographic choice sets both when accounting for ANA and when not considering ANA.