In a perfect market, employers will offer wage and bonus packages that will provide employees with just the right incentives to accomplish tasks they are given. But what if, as has been widely documented, employees are systematically over or under-confident regarding their abilities? If so, employers might be able to take advantage of these biases. For example, employers could offer overconfident employees low wages and very high bonuses tied to outstanding performance. Overconfident employees will accept the low wages because they assume (incorrectly) that they are very likely to get the big bonuses.
To test such predictions, the PIs will conduct an experiment in which participants are first assessed in terms of their self-confidence by comparing each participant's score on an aptitude test to the score the each participant predicted he or she would get. Subjects then participate in a simulated employment situation in a computer laboratory. The effect of over- or under-confidence is measured by comparing behavior when participants' probability of success in their task is determined by their test score, and when it depends on a randomly assigned number. This setup holds everything but the employee's self-assessed ability constant. The experiment will confirm if different self-confidence levels are indeed associated with different incentives, effort levels, and overall profits for employers and employees. Ultimately, it will set the stage for exploring the connection between self-confidence and demographic factors like gender and the implications of gender differences for the structure of work contracts.
