Underlying Contextual Effects Leading to over Consumption: Extremeness Aversion and Bundling

Obesity is now a global problem. Within the U.S., the rise in obesity has been largely driven by the increase in caloric consumption. By specifically observing the practices in the fast food industry, this dissertation examines two factors, portion sizes and bundling, to answer the question, what environmental factors in this context have promoted over consumption. With the first factor, portfolio of portion sizes, I demonstrate how extremeness aversion, the tendency for individuals to avoid the smallest and largest sizes, has gradually led consumers to choosing larger and larger portions. Historically research, as well as a consumption study included in this dissertation, demonstrates that choices of larger portions lead to greater end consumption. In regards to bundling (the common practice of offering an entrée, side item, and drink, often referred to as a "combo meal" or "value meal") this dissertation demonstrates that this practice induces people, who choose the bundled meal (on average) purchase larger quantities of side items and drinks. Though offering a combo meal induces consumers to be more price sensitive, those who choose the bundle, tend to be the least price sensitive individuals. Using Bayesian Estimation, this dissertation estimates the magnitude of these effects for each individual, controlling for any price effect on profits, choice and consumption. From the estimates, I produce a demand function for a fast food firm and run policy experiments. The policy experiments are not only used to understand which factors lead to the greatest caloric consumption, but also to evaluate possible policy actions (e.g. taxes) to reduce overall consumption. These experiments demonstrate that taxation of fast food does decrease consumption, however, at a great expense to firms and consumers. Taxes harm consumers because price increases through taxation reduce consumer surplus, and in this context, hurt the poor the most given the greater proportion of the poor consuming fast food. Thus I conclude, through policy simulations, that standards for portion sizes achieves the same or greater decrease in consumption as taxes with limited decrease in firm profits and no harm to consumers. Concluding this dissertation is an addendum on model comparison.