Second-best Congestion Tolling with a Heterogeneous Value of Time

Second-best Congestion Tolling with a Heterogeneous Value of Time


V van den Berg, E T Verhoef, VU University, NL


We analyse second-best congestion pricing in a bottleneck model with heterogeneity in the value of time. We find that the welfare with a pay-lane decreases with heterogeneity. This the converse conclusion of the literature, who used flow congestion.


In analysing the effects of transport policies it can be very important to adequately control for heterogeneity. Previous studies have concluded that ignoring heterogeneity can cause substantial biases in the calculated welfare effects of tolling. This paper analyses the welfare effects of (second-best) tolling, in a bottleneck model, with a continuous distribution of the value of time. Different from with homogeneity, under heterogeneity the generalised prices change when a first-best public toll is set: the low value of time drivers lose and the high value of time users gain. We find that in our bottleneck model, the average congestion externality decreases with heterogeneity. Hence, there is less to gain from public first-best tolling, and thus the welfare gain of this toll decreases with the degree of heterogeneity. The more heterogeneity there is, the lower the relative efficiency of a public pay-lane. This finding is in contrast to the previous literature: other studies, using static flow congestion, concluded that this relative efficiency increases with heterogeneity. With more heterogeneity, the relative efficiency of a private pay-lane is lower, whereas of a public time-invariant toll it is higher. Our results suggest that ignoring heterogeneity can not only bias the calculated welfare gains of policies, but also that the welfare gains of different tolling schemes are affected differently by heterogeneity. Moreover, how heterogeneity changes the welfare effects of policies can also depend on the type of congestion considered.


Association for European Transport