Introduction: Research has identified many factors associated with bicycling, but little is known on their relative influence for an individual's decision to bicycle or what led to the surge in bicycling during the COVID-19 pandemic in the U.S.
Methods: Our research leverages a sample of 6735 U.S. adults to identify key predictors and their relative influence on both increased bicycling during the pandemic and on whether an individual commutes by bicycle. LASSO regression models identified a reduced set of predictors for the outcomes of interest from 55 determinants included in the modeling.
Results: We find individual and environmental factors have a role in explaining the shift towards bicycling-with key differences in predictors for increased overall cycling during the pandemic compared to bicycle commuting.
Conclusions: Our findings add to the evidence base that policies can impact bicycling behavior. Specifically, increasing e-bike accessibility and limiting residential streets to local traffic are two policies that show promise for encouraging bicycling.
Keywords: Cycling determinants; Lasso regression; Public health; Transport policy.