Trapped by wisdom: how non-overestimating leads to underuse

This paper analyses time-consistency of decisions to visit the gym when individuals are asked to choose between their current fixed fee and a menu of three new variable fees, according to three different time commitments of permanency. We conduct a natural experiment on 290 members of a health-club, located in Barcelona from December 2009 to May 2010. Subjects are offered to switch from their former flat fee to a multi-part tariff one in June 2009. For the study period we collect for each subject: former fee, new fee (if changing), number of attendances per month during the study period, gender and age. We first look for consistency in their predecision behavior by comparing their flat fee with the one-day-entrance ticket option. Second, we measure consistency in their decision of changing or not to the new menu of fees according to their pre- decision number of attendances. Finally we observe post decision attendances in order to detect behavioral biases derived from the change in the structure of the contract. We find a relation between expected attendance estimation’s accuracy and level of consistency. We observe statistically significant behavioral bias for switchers that show a tendency to underuse their “free” units of their multi-part tariff.