A Regret-Matching Model for Analyzing the Effects of Switching Behavior to Teleworking

Abstract

With the recent development of information technology, teleworking has become remarkable as a solution for problems such as environmental issues, shortage of manpower for labor due to a low birthrate and decline in urban functions at the time of a catastrophic disaster. Both central and local governments show a positive attitude toward the spread of teleworking in Japan. Local governments have established a satellite office, and are promoting policies that positively recommend the introduction of teleworking. However, theory and techniques for forecasting how many people really accept teleworking when the working conditions for telework are not clear, still are not yet established. In this paper we focus on predicting the demand for the at-home employment type teleworking. We applied a game theoretical approach to describe the switching behaviour of a worker who chooses teleworking or ordinary commuting when he can choose either. We used the conventional approach for incorporating telework links as virtual links into transportation networks to analyze actions to choose teleworking. This allowed us to use the modal-choice model. However, our approach is different from the traditional modal-choice models in that we used the regret-matching scheme that corresponds to generic games. We showed that our approach is useful for predicting switching behavior between teleworking and ordinary commuting when a worker can choose. For example, one of our simulation results shows that the opportunity costs of each action converge to an equal level. Here we assume that the opportunity costs of teleworking are the losses in benefits from face-to-face contacts experienced during the ordinary working style. Furthermore, our approach is very flexible to the ill-defined cost function such as non-convex function, cost function with asymmetric Jacobian and non-differentiable function. These indicate that the proposed method has broader applicability in analysing complex system. © 2009, JAPAN SECTION OF THE REGIONAL SCIENCE ASSOCIATION INTERNATIONAL. All rights reserved.