The (un)compromise effect

The current study provides the first experimental test of the compromise effect, i.e. the tendency to choose middle options, in a naturally occurring setting. Simultaneously, I propose and evaluate a novel nudge intended to stimulate active choice - the (un)compromise effect - a compromise effect without an explicit middle option. 63,494 recipients of a mail fundraiser were randomly assigned to one of three sets of suggested donations: [$10, $50, $100, $__ ]; [$10, $50, $100, $250, $500, $__ ]; or [$10, $500, $__ ]. The results support both the compromise effect and the (un)compromise effect: extending the range increased the fraction donating $100 as well as the average donation - independent of whether the middle suggested donations were present or not. Hence, by only providing informative end points, organizations can affect decision-making and at the same time promote individuality through active choice. The results also shed light on why suggested alternatives affect choice in general: they reduce the cognitive cost of figuring out what actions are appropriate.