Do self-reported motivators really motivate higher performance?

Self-reported motivators, i.e., stated preferences for work incentives, have been examined extensively based on the assumption that they provide valuable information for the design of compensation systems. However, the extent to which these stated preferences match performance has been overlooked. I use a field experiment with sales representatives to examine whether a higher preference for an incentive leads to a greater performance effect when that incentive is introduced. Specifically, I examine whether ex-ante self-reported incentive preferences moderate incentive effects on ex-post objective performance. Using a setting in which sales representatives receive a fixed hourly rate, the between-subjects experimental manipulations add one of three incentive motivators: (a) money (i.e., monetary incentives in the form of cash bonuses), (b) feedback, or (c) recognition. My results show that being in the money condition (i.e., being eligible to win a cash bonus) leads to an increase in performance only for those who state a low preference for this incentive. Conversely, only those who state a high preference for feedback increase their performance when they are in the feedback condition. Finally, being in the recognition condition (i.e., being eligible to win an acknowledgement award) leads to an increase in performance regardless of the initially stated preference for this incentive.