BART Patronage Forecasts
Efficient, direct ridership estimation that accounts for effects of TOD, parking and transit access.

Ridership forecasting for prospective extensions of the BART rail transit system directly accounted for the effects of TOD and self-selection in the service area's mode choice. Statistical analysis of a full set of station-by-station data from the BART rapid transit and Caltrain commuter rail stations indicated, with a high degree of confidence, that transit ridership is directly and quantifiably related to:

• Population and employment within ½ mile of station
• Population in the station's overall catchment area
• Service type (BART vs. Commuter Rail)
• Peak period train frequency
• Number of station parking spaces
• Feeder bus service frequency

The resulting direct-demand forecasting model was used to predict ridership on proposed BART extensions through high-growth areas of the region. The forecasts indicated that ridership on the extensions would be:

• similar to performance of Western US light rail and commuter rail
• consistent with current trips per revenue mile of existing Northern California rail
• similar to actual ridership at stations with similar or lower train service and catchment-area activity
• 20% higher if station areas developed with transit-supportive land use, in keeping with current City policies, rather than the land use patterns common in other corridors