Killing the (coordination) moment: How ambiguity eliminates the restart effect in voluntary contribution mechanism experiments

The restart effect occurs in linear voluntary contribution mechanism (VCM) experiments when there is an upward pulse in contributions to the group account following a stoppage and then restarting of the VCM experiment. Although the restart effect is a well-known empirical regularity little research has been conducted regarding its causes. However, other scholars have noted that some kind of revision takes place during the stoppage that promotes cooperation. This research posits that certain common knowledge about when the stoppage occurs creates a “coordination moment” where group members simultaneously engage in revision and attempt to re-coordinate on a higher contribution level. Following Schelling’s description that such coordination comes from focal points that are both “prominent and conspicuous” I design a VCM experiment that reproduces the standard restart result and then compares those sessions to sessions where the exact pattern of stoppages is unknown. I find that this subtle manipulation of ambiguity (over the when stoppages will happen) eliminates the restart effect.