Highway Expansions, Tolls, and Congestion with Robert Krol

Episode Summary

Executive Summary: The episode argues that widening roads can ease congestion briefly but usually induces more driving, relocates activity, and restores gridlock. Guest Robert Kroll explains why economists favor congestion pricing: variable tolls better allocate scarce road space, reduce peak demand, and send useful signals for infrastructure planning. He also addresses implementation, privacy, public transit limits, and zoning.

Main Topics: Why road expansion does not solve congestion (Priority: 5/5): Kroll argues that adding lanes can raise speeds and throughput in the short run, but the lower cost of driving attracts more trips and activity back to the highway, bringing congestion back. Induced demand and behavioral responses (Priority: 5/5): The discussion covers how commuters shift travel times, public-transit users may switch back to cars, and households/businesses can relocate farther from core areas when roads get faster. Congestion pricing as the preferred policy (Priority: 5/5): Kroll explains the externality problem on free roads and advocates variable tolls that rise during peak periods and fall off-peak to smooth demand and improve efficiency. Implementation, technology, and privacy (Priority: 4/5): They discuss electronic tolling via transponders, license-plate billing, and the administrative and privacy challenges involved in road pricing systems. Public transit and fixed rail limits (Priority: 4/5): The guest argues that transit works best in dense, older cities and that fixed rail often struggles to reduce congestion in newer, spread-out metro areas. Urban planning, zoning, and land use (Priority: 3/5): The conversation explores transit-oriented development and zoning reform, with skepticism about top-down planning and emphasis on prices reflecting true social costs.

Key Arguments: Road expansion can improve traffic temporarily, but induced demand tends to restore congestion over time. When driving becomes faster and cheaper, people change behavior: they leave earlier/later, switch modes, live farther away, or change where they shop and work. Congestion is an externality: each extra car slows everyone else down, so unpriced roads are inefficiently overused. Variable tolls/peak-load pricing are the most direct way to manage congestion because they make drivers face the social cost of peak-hour travel. Tolls can reduce congestion without needing to remove many cars; even a modest reduction can improve speeds significantly. Electronic tolling technology makes congestion pricing feasible, though political acceptance and privacy remain obstacles. Public transit and fixed rail can help in some dense cities, but they generally do not eliminate congestion because they also face induced-demand effects and depend on urban form. Toll revenues and price signals can help planners identify where infrastructure is most valued and potentially reduce the need for costly expansions.

Data Points: Time horizon for congestion returning after added capacity: 3 to 5 years - Kroll cites the Los Angeles 405 example, where added capacity improved traffic at first but congestion returned after several years. Illustrative tolling reduction needed for meaningful speed gains: About 10% of cars removed - Kroll says even a small reduction in peak traffic can significantly improve travel speed. Typical congestion cost components: 4 categories - He lists time delay, fuel, business/shipping delays, and environmental costs as the main components used in congestion-cost estimates. Example carbon price used in cost estimates: $25 per ton of CO2 - Mentioned as an illustrative market price used when monetizing environmental costs from idling traffic. Commuter timing example: 1 hour - Used to illustrate how a driver may sleep in longer once new lanes reduce travel time. Locality example: Vancouver, B.C. bridge toll removal - Used to discuss how removing a flat toll increased congestion and why peak-only tolling might have been preferable. Illustrative historical toll booth payment: 25 cents - Example from the Garden State Parkway toll booths in New Jersey. Use of transit in older cities: New York cited as a successful dense-city example - Used to argue that fixed rail and public transit work best where density supports them.

Pivotal Quotes: "I think it's really hard to build your way out." — Robert Kroll: His direct answer to whether highway expansion can solve urban congestion. "What we really need to do is charge a toll. And the toll should vary depending on the level of congestion." — Robert Kroll: Core argument for congestion pricing as the efficient policy response. "A toll is actually an efficient tax. It's actually moving us in a direction of more efficient use rather than a less efficient use of resources." — Robert Kroll: Explains why congestion pricing differs from a standard distortionary tax.

Implications: For listeners and policymakers, the episode suggests that congestion pricing is more effective and fiscally smarter than endlessly expanding roads or rail. The big challenge is political acceptance, not technology.