Comment on “Naiveté, Projection Bias, and Habit Formation in Gym Attendance”

I show that a central result in Acland and Levy [Acland D, Levy MR (2015) Naiveté, projection bias, and habit formation in gym attendance. Management Sci. 61(1):146–160] is based on erroneously reported statistical estimates. In an experimental study, the authors provide financial incentives for regular attendance at a gym and report that these incentives induce a habit formation effect, resulting in sustained increased gym attendance after incentives are removed. This habit formation effect is not predicted by participants ex ante. The correct statistical estimates, however, do not confirm a habit formation effect. Participants’ ex ante expectations are thus correct in that they do not predict a habit formation effect, as there is none.

This paper was accepted by Uri Gneezy, behavioral economics.