Abstract
This study proposes a congestion pricing scheme using tradable bottleneck permits (TBP) and Pareto improving even if the revenue is not refunded to drivers. TBP scheme is originally one of the first-best time-varying pricing schemes, but does not always achieve a Pareto improvement when marginal utility of toll cost changes among drivers. This study aims to analyze the effects of TBP on departure time choice of drivers. In particular, we show the time dependent utility of drivers when they have different schedule flexibility and marginal utility of toll cost. The cases with and without TBP scheme are analytically compared. Then, we propose the TBP scheme which is Pareto improving without revenues refunded to drivers. We focus on a one-to-one network with a single bottleneck and employed a departure time choice model to discuss the case that there exists heterogeneity in schedule flexibility and marginal utility of toll cost. We assume two classes for two attributes as the heterogeneity respectively: “busy/free” and “rich/poor”, and formulated the drivers' utility changes caused by implementation of TBP. In this assumption, we show that a Pareto improvement is not achieved, where the utility of “busy-poor” group was decreased by the effect of TBP. We propose partial implementation of TBP as a scheme for a Pareto improvement. In this scheme, the bottleneck capacity is assigned to drivers with and without TBP, where the driver who has a bottleneck permit can pass through the bottleneck without congestion and a driver without it goes through congestion. As a result, we reveal conditions in the amount of the bottleneck permits which can satisfy a Pareto improvement. This study finally discusses the requirement for a Pareto improvement that was derived by the proportion of the amount of issued bottleneck permits, and the number of each class drivers.
