The bottleneck model: An assessment and interpretation

- Provides a consistent dynamic analysis to congestion while accounting for scheduling choices.
- Generates insights into how congestion develops that are surprising and often quite general.
- Illuminates distinct sources of inefficiency in addition to the main one of unnecessary queuing.
- Has been applied to real networks.
- Likely to be developed for practical applications by adding more realistic features.

The bottleneck model of congestion with endogenous scheduling has become a standard tool of transportation economics. It provides surprising insights about the time pattern of congestion, optimal pricing, and many distinct inefficiencies of unpriced equilibria including wrong departure order with heterogeneous preferences, wrong allocation of users across links of a network, and wrong order in which parking spaces are occupied. It illuminates the roles of travel-time reliability, traffic information, and extreme congestion (“hypercongestion”). It has been developed for use in practical network planning. Future use will probably emphasize greater realism, leading to more practical applications.