Impacts of Covid-19 mode shift on road traffic

TL;DR

This paper models how shifts away from transit and carpooling during Covid-19 can increase road congestion and travel times in US metro areas, highlighting the potential benefits of remote work.

Contribution

It introduces a model to quantify how changes in commuting behavior impact travel times, aiding policy decisions for managing congestion.

Findings

Transit and carpool shifts can significantly increase travel times in congested areas.

Remote work can offset travel time increases caused by mode shifts.

Model estimates show potential travel time increases of up to 20 minutes in major cities.

Abstract

This work considers the sensitivity of commute travel times in US metro areas due to potential changes in commute patterns, for example caused by events such as pandemics. Permanent shifts away from transit and carpooling can add vehicles to congested road networks, increasing travel times. Growth in the number of workers who avoid commuting and work from home instead can offset travel time increases. To estimate these potential impacts, 6-9 years of American Community Survey commute data for 118 metropolitan statistical areas are investigated. For 74 of the metro areas, the average commute travel time is shown to be explainable using only the number of passenger vehicles used for commuting. A universal Bureau of Public Roads model characterizes the sensitivity of each metro area with respect to additional vehicles. The resulting models are then used to determine the change in average travel time for each metro area in scenarios when 25\% or 50\% of transit and carpool users switch to single occupancy vehicles. Under a 25\% mode shift, areas such as San Francisco and New York that are already congested and have high transit ridership may experience round trip travel time increases of 12 minutes (New York) to 20 minutes (San Francisco), costing individual commuters \$1065 and \$1601 annually in lost time. The travel time increases and corresponding costs can be avoided with an increase in working from home. The main contribution of this work is to provide a model to quantify the potential increase in commute travel times under various behavior changes, that can aid policy making for more efficient commuting.