Mechanism Design for Personalized Health Care: An Experiment Motivating Healthy Habits

Register: bit.ly/3FTrJhg

Incentives for health behaviors are an increasingly important policy tool in both developed and developing countries, and there is widespread interest in improving their effectiveness. However, different contracts are likely to be more effective for different people. Mechanism design offers two strategies to improve contract effectiveness—tagging on observables (i.e., 3rd-degree price discrimination), and offering a menu of contract choices (i.e., 2nd-degree price discrimination)—but a key concern with both is that participants with private information might self-select into contracts that are favorable to the agent but less effective from the perspective of the principal. We adapt each of these strategies to customize incentive contracts for walking. Using a randomized controlled trial among more than 5,000 adults in urban India, we show that both mechanisms increase physical activity, leading to a 75% increase in steps walked relative to the effect of a one-size-fits-all benchmark. Moreover, we find that the concern that participants will self-select into less effective contracts is not only misplaced, but exactly backwards. Instead, a common force in health behavior settings—commitment motives—leads agents to prefer more effective contracts under both mechanisms. In particular, sophisticated time-inconsistent agents demand contracts that commit their future selves to walk more, bringing their preferences in partial alignment with the principal and improving the effectiveness of customization.

Ariel Zucker is an assistant professor in the Department of Economics at UC Santa Cruz. Her research studies policies to improve health and environmental conditions among underserved communities worldwide. Many of her projects focus on countering behavioral biases in personal decision making. Prior to arriving in Santa Cruz, Dr. Zucker did a postdoc at UC Berkeley ARE, and earned her Ph.D. in economics from MIT.