ABSTRACT
Obesity is a rapidly growing epidemic in the United States and a major public health challenge worldwide. To counteract this epidemic effectively, better understanding of its mechanisms are needed—we must understand not just what factors play a role, but how and why they matter. Most studies to date have focused on prices, technology, and the general availability of food. Less attention has been paid to the roles of social influence and the physiology of energy balance—despite growing evidence that both play important roles. In this paper, we present some initial findings from our analysis of two non-price mechanisms for obesity: the physiology of dieting, and socially influenced weight changes.
We show how the core equations governing the physiology of weight change can generate many of the known facts about diet and weight gain, including: the difficulty of maintaining a diet over a long period, high rates of recidivism after dieting, and substantial individual heterogeneity in the success of different types of diets. Using a new quantitative index of recidivist temptation, we develop a range of novel diets.
The notion that social norms are implicated in the obesity epidemic is not new. However, we show how a simple conformist social mechanism alone can drive a sharp increase in average weight. For initial weight distributions satisfying criteria identified here—and met by U.S. obesity data—a simple “Follow the Average” (FTA) weight adjustment rule generates increased mean weight. Indeed, the general FTA process, discussed mathematically below, can generate a rich variety of dynamics beyond obesity, including oscillatory behavior for which no conformist explanation has been considered.
We argue that integrative models adding such social and physiological mechanisms to economic ones will provide deeper explanations of the observed dynamics of obesity and a powerful array of policy interventions tailored to specific communities and individuals within them. The paper concludes with a sketch of one such model.