Understanding preferences for bicycling and bicycle infrastructure

Abstract

Using survey data collected in New Jersey, we analyze the frequency of bicycling and respondent perceptions of the safety of various bicycling facilities. Data was collected via a mixed-mode survey design, including intercepts, bicycle hangers, flyers in bicycle shops, and a Facebook advertisement targeted toward bicyclists in New Jersey (N = 1937). This provided us with a reasonable sample of respondents that included bicycle commuters and non-cyclists. Data on cycling frequency was collected for recreational and commute trips. Respondents ranked the relative safety of images of bicycle facilities, ranging from cycling on-street to off-road trails. We also collected attitudinal data on risk perceptions and world views linked to political perspectives. Our analysis suggested that feeling safer with on-street bicycle lanes and off-street bicycle paths is not associated with the frequency of bicycling, while feeling safer on-street in traffic is associated with the frequency of bicycling. In analyzing correlates associated with our images of bicycling infrastructure, we found those with more liberal/egalitarian world views prefer on-street bicycle lanes and off-street bicycle paths, while those with traditional community world views tend to not feel safe with on-street bicycle lanes and bicycling in traffic. Those who are risk takers also feel safer bicycling in traffic. Most other demographic controls in our models give us the expected results. Policy implications suggest that bicycle infrastructure will be less controversial when world views are more liberal/egalitarian, but that making all streets safer might be a useful approach for increasing the frequency of bicycling.

Keywords:

• Bicycling
• bike infrastructure
• risk perceptions
• political attitudes

Acknowledgements

James Sinclair contributed to this effort, as did Mark Walzer from the Bloustein Center for Survey Research. Thanks also to Kelcie Ralph and Mike Smart who provided useful questions for the survey and engaged in discussions over the research. All errors and omissions are the responsibility of the authors.

Authors’ contributions

R. Noland: Survey and research design, Supervision of analysis, Manuscript writing, idea for project. M. Laham: Conducted analytical work, contributed to manuscript writing, literature review. S. Wang: Conducted analytical work.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Updated statistics are available here: https://www1.nyc.gov/html/dot/html/bicyclists/bikestats.shtml, and show specific details on growth in the bicycle network in New York City.

2 In the study cyclists were defines as those who cycled more than weekly, were moderate transit users, and rarely used a car (Palm & Handy 2018).

3 Only 50% of households currently have landlines, and a greater fraction of young people are only reachable via cell-phone.

4 The data used for this study was collected by the Alan M. Voorhees Transportation Center (VTC) and the Bloustein Center for Survey Research (BCSR) in the Fall of 2016, between September 16^th and November 29th. The average time to fill out the survey on-line was about 10 minutes and our informed consent text indicated it would be about 15 minutes.

5 Intercept surveys were distributed in university towns such as Princeton and New Brunswick, shore communities such as Asbury Park and Belmar, and denser cities, such as Jersey City and Hoboken, and cities with transit-oriented development, such as Collingswood, Morristown and Metuchen. A full listing of locations is available in (Noland et al. 2016).

6 Our data collection protocol was approved by the Rutgers Institutional Review Board, Protocol#: 16-551 “New Jersey Bicycle and Pedestrian Resource Center”.

7 A prior test showed the advertisement to cell phone users was not effective.

8 Respondents who answered that they had not cycled in the last 12 months were not presented with the frequency questions. Those that answered the frequency question were still presented with “never” as a response (as they may have stated that they never commute, but they still cycle for other reasons).

9 We explored dropping the one variable with loadings below 0.4, “Many people I know think that using a bicycle for most trips is a good thing.” This resulted in no difference in interpretation of the factors. Given that these measures are relatively subjective, we decided to keep the loadings on this variable in our factor scores.

10 Population density was from the American Community Survey, 2015 five-year average. Residential locations were obtained for those respondents providing address information for the survey incentive drawing. Noland et al. (2016) provides a map of all locations.

11 We likely lost some respondents with more extreme views as evidenced by comments on the Facebook posting. For example: “When you started asking questions about my political stance on social issues I realized this survey was just a ruse. G'Bye! Figures, it was posted on a government site” and “Yeah, so…˙˙how did this go from a survey about bicycle usage to a survey assessing our views on socialism?”

Additional information

Funding

We would like to thank Charles Brown of the Bicycle-Pedestrian Resource Center at the Alan M. Voorhees Transportation Center for providing support for this analysis and for funding the collection of data.