Bonuses, goals, and instrumentality effects.

Three experimental studies showed that bonuses based on end-of-period determinations of standards led to the setting of more challenging goals but lower performance than a control condition in which bonuses were based on the achievement of self-set goals. Performance differences between the bonus and control conditions were not mediated by levels of self-set goals or goal commitment as predicted by goal theory. However, self-set goals and self-efficacy were significant predictors of performance within both the bonus and control conditions. Changes in performance under the end-of-period bonus condition in Study 3 were fully mediated by judgments of instrumentality. Participants in the end-of-period bonus condition were less certain of receiving a bonus, and this negatively affected their performance. Implications for the use of appraisal ratings to allocate bonuses and for the design of bonus schemes for management are discussed.