Examining the Cycle: How Perceived and Actual Bicycling Risk Influence Cycling Frequency, Roadway Design Preferences, and Support for Cycling Among Bay Area Residents

Abstract

This dissertation investigates the connection between perceived and actual bicycling risk, and how they both affect and are affected by one's attitudes, knowledge, behavior, and experiences. Understanding bicycling risk has gained importance as efforts by the U.S. Department of Transportation, the Environmental Protection Agency, the Centers for Disease Control & Prevention, and others have urged communities to increase cycling for its health, environmental, and social equity benefits. Research has identified numerous barriers to increased bicycling in the U.S., including topography, weather, and trip distance, but the barrier that appears most consistently between studies is the perceived hazard associated with cycling near motorists. Yet, little research has fully explored the concept of risk to understand its component parts, including how 1) various driver actions affect perceived and actual cycling risk, 2) reported crash statistics reflect perceived and actual risk, 3) roadway design preferences are affected by perceived risk, and 4) attitudes toward cycling and cycling risk--especially among drivers--influence support for bicycling in one's community. A deeper understanding of perceived and actual risk is critical for knowing how to address it, and, ultimately, to encourage more people to bicycle. To begin to answer these questions and demystify bicycling risk, this dissertation employs three main methods: focus groups, an online survey (n=463), and an analysis of reported crash data from the San Francisco Bay Area, one of the regions at the forefront of cycling efforts in the U.S.My findings confirm that perceived and actual cycling risk influence the decision to bicycle, but indicate that the causal pathways are more nuanced than previously understood. First, my data suggest that cyclists experience two types of roadway risk: pervasive risk in the form of near misses that occur frequently, and acute risk that occurs when a cyclist is struck--a less frequent, but more injurious incident. Both types--but particularly near misses--significantly affect perceived risk for cyclists and their family and friends, yet we lack systematic data on near misses and are therefore almost completely ignorant about the extent and effect of their occurrence. Routinely-collected reported crash data provide only limited insight into the type and extent of risk cyclists experience.Second, roadway design preferences are significantly related to perceived risk, and particularly important for attracting new cyclists. Surprisingly, drivers and cyclists both prefer roadway designs with separated space for bicyclists, particularly if barrier-separated, regardless of cycling frequency. Shared space designs are less popular among drivers and much less popular among cyclists, particularly for people who might consider cycling but do not currently do so: only a tiny fraction of potential cyclists feel comfortable sharing space with drivers on commercial streets. Third, perceived cycling risk extends beyond fear of danger for oneself, and is significantly related to support for cycling in one's community. Structural equation models of perceived cycling risk, attitudes, and behavior revealed that respondents are affected by their perceived risk as cyclists, but also as drivers sharing the roadway with cyclists they view as "scofflaws", and the risks they project onto other cyclists--particularly those cycling with children. This multi-pronged belief in cycling risk significantly negatively affects bicycling support, including support for new bicycle facilities and public funding to encourage cycling. Based on these findings, I propose a revised theoretical framework for conceptualizing cycling risk and its influences. I conclude the dissertation with policy recommendations for addressing perceived risk.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 26 citations exist upstream.