Feasibility Study for a Bicycle Sharing Program in Sacramento, California

This study presents a methodology for analyzing the feasibility of bicycle sharing programs based on U.S. experiences. While past studies have had to rely on observations from European systems, the recent emergence of several large-scale programs in the U.S. creates an important opportunity to explore the determinants of bike-share use in a more relevant context. This understanding will lead to more effective strategies for promoting the bike-share mode in pursuit of reduced congestion, enhanced air quality, and improved public health. This project analyzed recent data from a bicycle sharing program in Minneapolis (the “Nice Ride” program) and applied the findings to Sacramento, where a program is currently being considered. A regression analysis of Nice Ride rentals was conducted to identify the determinants of monthly rentals and to establish a predictive model. These findings were then evaluated in Sacramento to recommend a program service area and predict ridership. First, a weighted sum raster analysis was performed to identify suitable locations for a program and to recommend service areas. Second, the coefficients of the Nice Ride regression model were applied to characteristics of the recommended service areas to predict ridership. Estimated ridership ranged between 26,864 and 28,266 rentals per month. These estimates can be used to predict program revenues and facilitate effective decision-making. The study also found that underlying differences between Minneapolis and Sacramento played a key role in the results, indicating that this methodology should be repeated for pairs of more similar cities as data from more U.S. programs become available.