How Past Experiences Shape Future (Prosocial) Behaviors: A Large-Scale Natural Field Experiment

Abstract

Future behaviors are shaped by the content of past experiences, but scholars have assumed that the order of past experiences should not matter at all. Using a large-scale natural field experiment involving 14,383 volunteer crisis counselors, we examine how the content and order of past experiences causally influence future behavior – namely, whether individuals continue volunteering or quit. Volunteer crisis counselors were repeatedly and randomly assigned to perform 1,976,649 time-stamped prosocial behaviors that were either harder prosocial behaviors (suicide conversations) or easier prosocial behaviors (non-suicide conversations). We find that the content of past experiences mattered: Harder (versus easier) behaviors caused volunteers to become more likely to quit. However, we find that order mattered up to 99 times more than content: Harder behaviors caused disproportionately more quitting if they came in long streaks or at the end. Two aspects of our investigation are especially noteworthy. First, our investigation is the largest field experiment on prosocial behavior ever conducted in terms of the number of randomizations. For example, our investigation is 49,000 times larger than a classic study about helping behavior by Darley and Batson (1973). Second, our results suggest that if leaders implemented a simple “reordering intervention” (i.e., never assigning hard-behavior streaks or hard-behavior ends to volunteers), it would double prosocial behavior and save lives. Strikingly, the reordering intervention’s expected percentage point impact is 449% larger than the impact of any previous intervention mentioned in a prominent review of leading interventions by Benartzi et al. (2017).