Date Available


Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation


Graduate School


Public Policy and Administration

First Advisor

Dr. J.S. Butler


Despite the efforts of policymakers, medical professionals, and other stakeholders, obesity and related health problems show no signs of receding from their record-high rates. Public policy has largely taken the form of consumer advice, (e.g., USDA’s Dietary Guidelines). Since consumers bear most of the costs associated with their obesity, the goal of obesity prevention appears to be incentive-compatible, prima facie. That is, there is no a priori case for much further policy intervention unless existing advice is deficient or consumers’ exhibit systematically poor decision-making.

My review of the literature shows that scholars have long conveyed a consistent narrative regarding our scientific understanding of obesity—one which emphasizes the apparent behavioral ‘anomalies’ of consumers. Unacknowledged by most investigators is a body of scientific literature which, if valid, severely undermines the predictive and explanatory power of most accepted models of obesity. That is, credible institutions may prevent consumers from discovering obesity-alleviating diets because nutrition authorities and policymakers have condoned only one approach to healthy eating for several decades. I advance a theory of obesity rooted in Shacklean uncertainty about the operationally relevant characteristics upon which consumers should base their decisions. I relax the standard assumption that consumers exhibit open satiation isocalorically.

To measure consumer preferences in the pre-obesity epidemic era, I perform a content analysis of American cookbooks from the early 18th to the middle of the 20th century. Results show that, if anything, past Americans preferred food that would be considered worse—even further from the USDA ideal.

Using USDA survey data (1994-1996), I construct a finite mixture model to analyze Americans’ consumption patterns, preferences, knowledge, and beliefs about diet and health. My estimation shows that Americans were aware of the tenets of healthy eating at that time and consumption patterns were broadly consistent with the strength of these beliefs and preferences. I argue that economic theory would not predict rates of obesity doubling in the interim, given consumers’ continued exposure to information and their experiences. I hypothesize that this is evidence for a significant discrepancy between institutionalized advice and advice which is truly effective— a form of uncertainty. I use the extant scientific literature to show the ways in which healthy decision-making is constrained relative to what it would be under counterfactual policy arrangements. I do not reject the null hypothesis that consumers would have been at least as well off with no policy at all. I use the same dataset to replicate Butler’s (1982) work which models the relationship between social stigma and the decision-making of consumers who receive food stamps.

Digital Object Identifier (DOI)