Discussion paper

DP17425 The Value of Choice - Evidence from an Incentivized Survey Experiment

Do people have a preference for making choices themselves, or do they prefer to choose a preselected alternative? If consumers value choice, recommender systems which facilitate choices might trigger consumers not to choose the recommendation - even when the other alternatives are less preferred. We conduct an incentivized survey experiment with a large sample from the German population, where participants choose between three lotteries. In the main treatment, participants make a choice between a preselected lottery and a two-element choice set, from which they then make an additional choice. We find that participants’ choices exhibit a bias towards the preselected alternative, and estimating a structural model reveals that the mean willingness to pay to make an additional choice is negative. Nevertheless, about 41% of the sample are estimated to have a positive value of choice. We show that measurable individual characteristics correlate with the preference for choice. Linking choices to the Big Five personality traits reveals that the preference for the preselected alternative increases in Openness. Moreover, we link participants’ preferences for choice to their political attitudes, showing that right-wing participants are more likely to prefer the preselected alternative than center-left participants.


Grüner, H and L Rohde (eds) (2022), “DP17425 The Value of Choice - Evidence from an Incentivized Survey Experiment”, CEPR Press Discussion Paper No. 17425. https://cepr.org/publications/dp17425