Choice, deferral, and consistency

We report on two novel choice experiments with real goods where subjects in one treatment are forced to choose, as is the norm in economic experiments, while in the other they are not but can instead incur a small cost to defer choice. Using a variety of measures, we find that the active choices (i.e., those that exclude the deferral outside option) of subjects in the nonforced‐choice treatment are generally more consistent. We also find that the combined deferral and active‐choice behavior of subjects in that treatment is explained better by a model of dominant choice with incomplete preferences than it is by rational choice. Our results suggest that nonforced‐choice experiments and models can be helpful in separating people's rational, hesitant/not‐yet‐rational and genuinely irrational behavior, and can potentially offer important new insights in revealed preference analysis.