Skip to content

Case study: pricing the Baltimore Beltway (I-695)

Model version

This study was run with an earlier version of the chain (before the re-estimated models documented on this site). It shows what the platform can do; the numbers will be updated with the current models.

Question

Who would pay a toll on I-695, who would change route or stop using the Beltway, and how would the benefits and costs be distributed across income groups?

Designs tested

  • Uniform tolls on the Beltway.
  • An area-based scheme.
  • An income-based discount (50 % for households under $50,000), applied person by person.
  • A fixed charge per entry, to test the mechanism.

Each design was run in MATSim with the same plans and seeds as the base, and the results were broken down by person: stayers, leavers and newcomers on the Beltway, travel time and toll paid, and consumer surplus by income decile.

What the platform showed

  • Beltway use is spread evenly across incomes, so a uniform toll takes a larger share of income from lower-income households: the toll is regressive.
  • Every design reduced net welfare in the tests: the users' loss in consumer surplus exceeded the toll revenue.
  • The income-based discount was the only lever that reduced the regressivity, at a cost in revenue.
  • Select-link analysis traced diverted Beltway trips to other routes, such as I-83 and arterials like MD-140.

Outputs produced

  • welfare and revenue tables by design and income decile;
  • maps of toll incidence by home location;
  • hourly link-volume animations;
  • diversion paths of leavers.