An agent based model of stress in the workplace

Abstract

This study introduces a simple agent-based model of stress in the workplace. Agents are described by time allocation to a base job, a special project, and time spent resting. Agent training is the inaccurate imitation of high productivity colleagues. The relative worth of the base job and special project vary stochastically, representing shifting management priorities. Stress is accumulated by working long days and has a negative impact on productivity. The impact of covert drug use, implemented as abnormally high stress tolerance, is investigated as is a form of proactive management encouraging workers to work longer hours. It is found that choosing the training cohort for low-performing and new workers to be other than the highest performers reduces overall productivity but also ameliorates the impact of having drug users training non-drug users toward unrealistic performance levels. Proactive management is found to have a small impact on productivity but sharply increases the number of low-performance related firings.