Association for Public Policy Analysis & Management

As performance management arrangements in the public sector proliferate, performance expectations will vary in terms of practicality. Negative feedback in this context is likely. Under such conditions, it is plausible that the same motivations that have proven to attract individuals to careers in public or nonprofit service (versus the private sector) will be diminished. Yet, while there is well-­established empirical evidence that individuals who are more pro­socially motivated and identify with altruistic or social organizational missions are more likely to self­-select into careers in the public or nonprofit sectors (e.g., Andersen et al, 2013; Perry et al, 2010), little to no evidence exists as to how durable those motivations might be in relation to altruistic work effort—especially in a context of negative feedback.

Our primary research question, then, is do employees put forth work effort increasingly toward their own self-­interest (rather than the purposive goals of the organization) when faced with negative feedback?

For this experiment, we recruit workers from the United States from Amazon’s Mechanical Turk (MTurk). We use a “real­-effort” experiment—an experimental approach that is common in economics (e.g., Tonin & Vlassopoulos, 2010)—to test these propositions. Subjects complete survey questions that measure PSM, self­-determination, demographic information, and political ideology. We then provide a battery of organizational missions and ask subjects to rank the missions by salience and valence.

Subjects then perform a real­-effort task for a set amount of time. The real-­effort task is a Simple Reaction Time Task (SRTT). Once the subject has completed the SRTT for the first time, they are provided their performance. They are then offered the choice to repeat the SRTT or to submit their work to complete the entire experiment. If they choose to submit, they are remunerated for their participation. However, with the offer to repeat, the subject is randomly assigned one of the five previously ranked charities. The subject is prompted that they will be given $10 if the subject repeats the task and improves their performance to one of three randomly assigned performance expectations. If the subject fails to meet the expected performance, they are offered the choice to exit the survey or “play again.” The subjects are given the opportunity to change the proportion of the $10 that benefits the charity in relation to one’s self. This is repeated to infinity, or until the subject achieves the expected performance or chooses to exit.

Expected results for this research are that subjects with high PSM will be more likely to repeat a real­-effort exercise on behalf of charity. At the same time, mission support will be an important predictor to work effort and the extent to which the proportion of reward will favor the charity (rather than one’s self). Finally, the more unrealistic it becomes to attain a given goal, the more a subject will choose to work in their own self-­interest than on behalf of a charity. However, this crowding­-out effect will be attenuated by the subject’s PSM.