Don?t pay the highly motivated too much

One remaining puzzle in the literature on intrinsic motivation and extrinsic incentives is the conflict between two oft-cited rules: "pay enough or don't pay at all"and "pay, but not too much". There is some evidence to suggest differing impact of incentives, depending on level of initial motivation. Our lab-in-the-field experiment, on a heterogeneous sample of the general public, allows us to test our intuitive prediction: that crowding out will be less (more) likely when motivation is low (high) to begin with. We use a real effort task and a within -and between-subject design to measure motivation before, during and after an incentive is applied. On average, we find support for "pay, but not too much", as a low power monetary incentive is as effective as a high power monetary incentive when applied. However, high power monetary incentives strongly crowd out motivation for the high types. Monetary incentives (low or high) do not crowd out motivation in low motivation types. A charity incentive does not increase performance in the real effort task, but also has no crowding out effect. We show the importance of accounting for initial intrinsic motivation, and that the highly motivated are more susceptible to crowding out.