Absence, overtime, and reliability relationships in transit workforce planning

Abstract

The optimal size of a transit operating workforce is based on determining the appropriate amount of overtime requested of the workforce. Overtime is typically used to fill in for absent operators if no extra operators are available to do the work on regular time. Relying more heavily on overtime has two risks. First no operator may be willing and available to work overtime when it is needed, resulting in missed service and hence poor service quality. Second operators may be absent more, knowing that they can readily obtain overtime work at asignificant wage premium to replace lost regular wages. This paper examines these interrelationships between overtime, absence, and service reliability which are critical to overall workforce planning. In order to study the relationship between overtime and absence, a disaggregate model of absence is developed as a dynamic form of motivated behavior, a problem in time allocation across activities. The model is used to test the hypothesis that widespread availability of overtime may induce absence. Panel data of surface transit operators from the Massachusetts Bay Transportation Authority (MBTA) were used with the tobit model. The results suggest that absence is more a habit than the result of a decision process based on past overtime worked. The relationship between overtime and reliability is studied using aggregate data from the MBTA bus system. The results show a strong linear relationship which makes it possible to include reliability constraints in the strategic problem by setting an upper limit on the amount of overtime that can be planned for that period.