Methodology

A regression model of dwell time by fare payment type was estimated. For the model, Metro provided Universal Farebox System (UFS) and Automatic Passenger Counter (APC) data from March 3 to March 16, 2014. Because prior work suggests that limited-stop services have different dwell times than local services, two contrasting Metro bus lines were selected for analysis: a limited-service, low ridership neighborhood route (Local Route 120) and a high volume, high frequency rapid route (Rapid Route 720).

After throwing out outliers, the sample size consisted of 540,407 fare payment records and 99,453 dwell time records across 342 operators and 187 vehicles. Linear regressions analysis was used to statistically control other factors known to influence dwell times – including number of passengers boarding and alighting, wheelchairs, bicycles, bus type and configuration, and service (local, express, etc.).

Findings

• Statistically significant coefficients show that smart card payments contribute roughly two seconds per person to dwell time, while cash or other media contribute about four seconds. If 100,000 current customers (or about 20 percent of current non-TAP paying customers) were to switch to TAP cards, Metro buses would spend about 56 fewer hours per day waiting at bus stops.
• The study also found that the effect of TAP card usage on dwell times appears to erode appreciably in crowded conditions, meaning there are likely more cost-effective ways to reduce dwell times – such as all-door boarding at high-volume stops and stations.