Factors associated with commute dissatisfaction and commute time dissonance: A sample selection modeling approach

Abstract

This study identifies determinants of commute time dissonance between actual and desired commute times and commute satisfaction to clarify the role of commute time dissonance in determining commute satisfaction. We hypothesize that (1) commute time dissonance is not a simple determinants of commute satisfaction, but a complicated measure affected by various factors, and (2) commute time dissonance could be a mediator accounting for the relation between explanatory variables and commute satisfaction. Two specific measures (gap and ratio) of commute time dissonance are newly defined. Due to non-randomly selected cases in our sample (only dissatisfied commuters reported their desired commute time in the survey), we employ a sample selection (SS) modeling approach, incorporating socio-economic traits, travel characteristics, lifestyle, and perceptions on neighborhoods as explanatory variables. The result shows that two parts of the SS model—selection (commute satisfaction) and outcome (commute time dissonance) models—share key variables, including commute characteristics, lifestyle, and perceptions on neighborhood. It implies that commute time dissonance can play a role as a mediator in explaining commute satisfaction; thus, further statistical controls to correctly estimate impacts of determinants on commute satisfaction are required. We believe that our findings and implications are worthwhile to deepen the understanding of travel satisfaction and the role of travel time dissonance in the context of commuting.