14 January 2014Economists recognize that monetary incentives can backfire through the crowding-out of moral and social motivations leading to an overall decrease of the desired behavior. We implement a field experiment where participants are asked to fill a questionnaire on pro-environmental behaviors under different incentive schemes, either with no monetary incentive (control) or with low or high monetary incentive directed either to the respondents or to an environmental cause. We investigate whether (i) there is a significant crowding-out effect, (ii) directing monetary incentive to the cause rather than to the respondents reduces the overall impact of a crowding-out effect, and (iii) offering the choice regarding the money recipient affects participation. Except for a high monetary incentive where the respondent chooses himself the end-recipient, we show that monetary rewards directed either at the individual or at the cause actually harms intrinsic motivations, but not to the same extent. We formalize our results building on an adaptation of an original model by Bolle and Otto (2010) and introduce agents heterogeneity in terms of intrinsic motivation. This heterogeneity has key implications for the understanding of thecrowding-outeffect. Several policy recommendations regarding the use of market-based instruments are drawn.