Power harassment is the act of exceeding the proper bounds of authority in the workplace in order to inflict psychological or physical pain, but an objective definition of the proper bounds is difficult. A study was therefore made to see whether the proper bounds could be determined from a quantitative analysis of court decisions in cases of power harassment. The study was performed by extracting the actions of workers and their superiors from court cases, encoding them, and analyzing them by the standardized error root mean square method of the MT system, taking actions that had only a small effect on the other party to constitute the unit space. When the mean value of the worker's standardized error root mean square was plotted on the X-axis and the mean value of the superior's standardized error root mean square was plotted on the Y-axis, the plot showed agreement with the court results: cases in which only small compensation was awarded fell close to the balance line (y=x) that indicated the worker's actions and the superior's actions had about equal effect on the other party. This indicated the possibility of determining whether the superior's actions exceeded the proper bounds of authority by comparing the standardized error root mean square values of the superior's and worker's actions.