Cycling and the built environment, a US perspective

This disaggregate cross-sectional study uses primary data on the cycling behavior of 608 randomly sampled respondents in urbanized King County, Washington, and objective parcel-level GIS measures of land use and infrastructure conditions. Binary logit model findings provide new insights on who bicycles, and on perceived and actual built environmental conditions associated with the likelihood of cycling in neighborhoods, controlling for socio-demographic variables. A high 21% of the respondents report cycling at least once a week in their neighborhood, more often for recreation or exercise than for transportation. Cycling is more popular among male, younger adults, transit users, and those who are physically active and in good health. Both perceived and objective environmental conditions contribute to the likelihood of cycling. Proximity to trails and the presence of agglomerations of offices, clinics/hospitals, and fast food restaurants, measured objectively, are significant environmental variables. Previously researched correlates of cycling, such as the presence of bicycle lanes, traffic speed and volume, slope, block size, and the presence of parks, are found insignificant when objectively measured. A non-linear relationship is found between the odds of cycling and the perception of traffic problems and automobile-oriented facilities. Overall, cycling is only moderately associated with the neighborhood environment. It appears to be an individual choice that is independent from environmental support. This finding likely reflects the limited bicycle infrastructure in the sample frame-an unfortunate condition found in most US metropolitan regions. Policy and intervention programs could increase cycling by improving both actual and perceived environmental conditions. (c) 2005 Elsevier Ltd. All rights reserved.