Bicycle Share Systems: A Predictor of Crime?

Abstract

This research examined the correlation between selected common street crimes and potential increases in human activity caused by bikesharing ridership in Boston and Cambridge, MA. Given the increasing domestic and global popularity of such systems, evidenced by their growing global trends of implementation, it is worth exploring their potential effects on other aspects of urban quality of life, such as rates of crime, and to explore using these data to improve such systems’ effectiveness and sustainability. This study focused on two specific research questions: 1) Does an increase in the ridership of a Bicycle Share System (BSS) reduce incidents of assaults, robberies, and larcenies in cities? 2) To what degree does ridership correlate with assaults, robberies, and larceny incidents when compared to correlations with other urban demographic factors? I hypothesized that Boston and Cambridge neighborhoods with a higher ridership of BSS have lower incidents of robbery, assault, and larceny, and that ridership correlations with robbery, assault, and larceny incidents are on par with, or stronger than, correlations to other demographic factors. Using various statistical and software applications, I analyzed BSS ridership against several relevant crime variables, and included other demographic, economic and geographic factors that the FBI considers as recurring factors associated with variation in crime rates. The results showed a positive correlation between BSS ridership and nearby crimes, indicating that when ridership increased, nearby instances of selected crimes increased as well. The existence of such a correlation is not an indication of causation; instead, it reflects the fact that there are other relevant demographic variables shared that may be causing both studied variables to increase, although the effect of such variables varied across crime and location. For example, the presence of retail and other businesses was statistically significant, while other factors such as median home value and education level had varying effects across crime types and locations. These results ultimately reflect the multifaceted and complicated nature of crime; therefore, the intersection of BSS activity with municipal criminal trends is one that merits further study given its potential for improving urban built environments to deter crime.