Most people experience the feeling of mental fatigue on a daily basis. Previous research shows that mental fatigue impacts information processing and decision making. However, the proximal causes of mental fatigue are not yet well understood. In this research, we test the opportunity cost model of mental fatigue, which proposes that people become more fatigued when the next-best alternative to the current task is higher in value. In four preregistered experiments (total N = 430), participants repeatedly reported their current level of fatigue and chose to perform a paid labor task vs an unpaid leisure task. In Study 1, all participants were offered the same labor/leisure choice. In Studies 2 and 3, we manipulated the opportunity costs of a labor task by varying the value of an alternative leisure task. In Study 4, we manipulated the opportunity costs of a labor task by varying the value of that labor task. In all studies, we found that people were more likely to choose for leisure as they became more fatigued. In Studies 2–4, we did not find that the manipulated leisure value influenced the amount of fatigue participants experienced nor the likelihood to choose for leisure. However, in exploratory analyses, in all studies, we found that participants who reported to value the leisure task more, got more fatigued during labor and less fatigued during leisure. Collectively, these results provide cautious support for the opportunity cost model, but they also show that cost-benefit analyses relating to labor and leisure tasks are fleeting.

In this research, we attempt to understand a common real-life labor/leisure decision, ie, to perform cognitive work or to interact with one’s smartphone. In an ecologically valid experiment, participants (N= 112) could freely switch back and forth between a doing a 2-back task and interacting with their own smartphone. We manipulated the value of the 2-back task (by varying the value of monetary rewards; within-subjects) and of the smartphone (by switching on and off airplane mode; within-subjects) while we recorded incoming notifications, such as text messages. Our study produced three main findings: 1) the current value of the smartphone did not increase our statistical model’s ability to predict switches from labor to leisure when the current task value was also taken into account; 2) however, participants reacted strongly to naturally incoming notifications, which were the strongest predictor of labor-to-leisure switches; 3) there was no evidence that taking into account individual differences (in the value assigned to labor and leisure) improved the model’s ability to predict labor-leisure switches. In sum, using a situated approach to studying labor/leisure decisions, our findings highlight the importance of high task motivation, as well as the temporary distractive potential of smartphone notifications, when people face the challenge to stay focused on their productive tasks.