The equity effects of congestion pricing have come into focus as a potential hindrance to its acceptability in implementation around the world. Both theoretical arguments and empirical evidence about how the burden of the toll should fall on demographic groups have been mixed, depending on a variety of contextual factors such as automobile access, work schedule, flexibility, and spatial distribution of activities. Yet the evidence so far for implemented congestion pricing systems has not examined explicitly the role of such contextual factors. With the congestion pricing trial in Stockholm, Sweden, as a case study, this paper uses structural equation modeling to estimate a model of the role of age, income, and gender as independent variables; contextual factors as endogenous variables; and the change in automobile trips after congestion pricing as dependent variables. The findings indicated that gender was a significant factor in terms of the total effects and that three of the four contextual variables-access to a car, possession of a longterm transit pass, and having a workplace on the same side of the cordon as home-played significant mediating roles in the indirect effects of the demographic variables on trips by automobile. Moreover, it was found that gender was significant only when the contextual factors were included in the model. This finding suggested that a significant variable could be missed when such factors were omitted.
2012. no 2297, 29-37 p.