Publication Detail

Cruising and On-Street Parking Pricing: A Difference-in-Difference Analysis of Measured Parking Search Time and Distance in San Francisco

UCD-ITS-RP-18-23

Journal Article

Urban Land Use and Transportation Center, 3 Revolutions Future Mobility Program

Suggested Citation:
Alemi, Farzad, Caroline J. Rodier, Christiana Drake (2018) Cruising and On-Street Parking Pricing: A Difference-in-Difference Analysis of Measured Parking Search Time and Distance in San Francisco. Transportation Research Part A 111, 187 - 198

When on-street parking is scarce, the cost of parking includes the extra time and fuel spent searching for a parking space (or cruising). Cruising also unnecessarily contributes to local congestion, vehicle emissions, air pollution, and climate change. The theoretical literature shows that these social costs can be reduced, or even eliminated, if high-quality information on the demand for and supply of parking is used to set parking prices at optimal levels. Not surprisingly, cities plagued by parking shortages, congested streets, and limited financial resources are interested in parking policies that reduce cruising and improve the efficient use of their existing parking and roadway infrastructure. The current study sheds light on the effect of the San Francisco parking pricing program (known as SFpark) on curbside parking search time and distance in urban neighborhoods on non-commuter parking. The study differs from previous empirical evaluations of similar parking pricing programs in its use of direct field measurements of parking search time and distance, rather than simulated data or proxy variables, such as parking availability. We use generalized mixed effect difference-in-difference models with data collected before and after the implementation of SFpark in both treatment and control areas to estimates effects of the San Francisco smart parking project, most importantly the demand responsive parking pricing scheme. The models control for time effects by using data from a separate control area, as opposed to using variables such as block face parking price and employment. The results suggest a significant reduction in average parking search time and distance due to SFpark. Average parking search time and distance declines by approximately 15% and 12%, respectively, from the control to the treatment areas.

Keywords: Cruising; Parking pricing; Parking search time; Parking search distance; Difference-in-difference; SFpark