If a previously unpaid activity (e.g. donating blood) is paid, then we often observe that this activity is reduced. In this paper, it is hypothesized that the price offered is taken as a proxy for the “value” of the activity. Depending on how the actor valued the activity previously, crowding-out or crowding-in is implied, an effect with or without persistence after stopping the payment. The model can be adapted to a number of similar situations, including those where a high price signals high costs instead of high values. Our “naïve” explanation is confronted with Bènabou and Tirole's (2003) Principal-Agent model. A questionnaire study supports our basic hypothesis as well as some of the derived consequences, and contradicts Bènabou and Tirole's model.