Skip to:


Supporting decision making for sustainable weight loss


  • Associate Professor, Medicine (CHP/PCOR)

This project explores fundamental questions about habit formation and the role of nudges and other behavioral economic devices. It considers these questions in the health-relevant context of generating exercise habits. The study develops behavioral economic theory and links it outcomes from our large randomized study of exercise commitment contracts and nudges.

The researchers analyzed follow-up data on more than 4,000 individuals seeking to make exercise commitments who were randomized to be nudged towards longer (20 weeks) vs. shorter (8 weeks) exercise commitment contracts. They found evidence that the original nudge helped to induce habit formation and the awareness of the need for ongoing commitment in a subgroup of individuals.

They described a multi-period hyperbolic discounting model of exercise and non-exercise consumption that accounts for nudges. The model includes the ability to form consumption habits - deriving greater utility from subsequent consumption as a function of past consumption. They analytically solved for optimal multi-period consumption paths that are stable from all periods' perspectives and characterize the comparative statics of the model across a range of utility parameters. The analyses yield intriguing predictions about the relationship of exercise patterns and the effects of nudges on consumption and utility as functions of how hyperbolic a discounter a person is. These predictions are broadly consistent with the result of our randomized empirical study and allow the researchers to address the welfare consequences of nudges.

Previously, the researchers analyzed a pilot sample of baseline information (n=700) which was published as an NBER working paper and was picked up by the New York Times Freakonomic blog and by VoxEU (an online outlet describing cutting edge economics work). Two summers ago they worked with an undergraduate from MIT to analyze follow-up data through 14 weeks. They were invited to attend and will be presenting their work at the Penn CMU Roybal Symposium on Behavioral Economics and Health (organized as part of a P30 funded by NIA). The researchers had their abstract on this topic featured as a plenary session/top-ranked abstract in the 2011 Society for Medical Decision Making conference whose theme was behavioral economics. Follow-up work was an invited presentation at both the University of Oslo Health Economics seminar and the combined Oslo/Bergin Health Economics workshop. Also, follow-up and theory work were presented at the 2013 SMDM Annual meeting.

This project has yielded two publications: “The behavioural economics of exercise habits” and “Committing to Exercise: Contract Design for Virtuous Habit Formation. National Bureau of Economic Research” (NBER Working Paper). This project also led to the receipt of another NIA supported grant by Stanford’s Center on the Demography and Economics of Health and Aging, entitled “Helping Older Adults to Initiate Exercise Habits”. They are currently finalizing the write-up of the analysis with full follow-up data and new theory and expect to produce another NBER working paper and a separate manuscript for submission to top-tier journal.