About the Bike Lab

The Bike Lab began as a field research project which formed part of my thesis work in the Urban Studies program at UC Berkeley. My current interests are in the causal relationships between cycling advocacy, bike facilities, utility cycling rates, and neighborhood change.

Having completed the program, I continue to research these areas, with a particular focus on the dynamics of cycling infrastructure projects in disadvantaged communities, and the different cultural relationships those communities have to the bicycle.

I am now available for consulting on these matters, as founder and principal of Totally Doable Consulting.

The project’s name, and some of its premises, were inspired by Steve Zavestoski (University of San Francisco), co-editor of “Incomplete Streets.” It is also heavily influenced by the work of the bicicultures movement led by authors such as Adonia Lugo, Melody Hoffman, and John Stehlin.

Research Premises

  • Existing data on cycling is inadequate.
    Cycling rates are generally reported from American Community Survey (ACS) data, which has the advantage of being regularly collected across the country, but the disadvantage of representing a very narrow view of utility cycling (limited to work commute only), and entirely neglecting social and recreational cycling. Cycling facility data are typically reported only in raw mileage, which obscures the quality and utility of those facilities. Bike Lab seeks to collect relevant data to better investigate the causes and effects of utility cycling in cities.
  • Non-commute utility cycling has increased substantially in the past decade.
    Experientially, urban utility cycling has increased in the Bay Area more rapidly than the reported ACS mode share for cycling. We look to collect additional data to investigate this trend and consider its implications for advocacy, infrastructure, and neighborhood change.
  • Individuals make mode share choices based on complex criteria.
    Cycling facilities have been shown to be correlated with cycling mode share, but the relationship is far from universal. Utility cycling can exist at high rates in places with limited designated infrastructure, and some places with extensive infrastructure see low utility cycling rates. My hypothesis is that the most predictive factors for utility cycling are residential density and cultural acceptability, but that many other factors are in play. In particular, low-income communities may have high utility cycling rates because of low car ownership rates; more affluent members of those communities are less likely to be utility cyclists. In gentrified urban neighborhoods the opposite may be true; more affluent members may view the bicycle as a status symbol of the new urbanism and may be more likely be utility cyclists.
  • Utility cyclists choose routes primarily based on utility.
    Jennifer Dill (Portland State University) has done significant work on cyclist route choice. Her work fits with my intuitive sense that route designations, and separation from cars, are relatively minor factors in utility cyclist route choices. Designated routes may achieve heavy utility usage if routes are well-located, and provide facilities which utility cyclists value (limited stops, relatively direct routes between desirable destinations, limited hills). Note that mode choice and route choice are investigated as separate concepts.
  • Appropriate infrastructure design is context-dependent.
    Cycling and pedestrian-oriented road projects have two foundational goals: reduce auto trips, and increase safety. Northern European cities such as Amsterdam and Copenhagen are often referenced as aspirational ideals for cycling infrastructure. However, infrastructure always must be installed in context, and the European models cannot be broadly implemented in most U.S. cities due to fundamental differences in the built environments. NACTO guidelines provide models for street and lane widths for road facilities, but a facility that is successful in one location may not be successful in another because of that context.
  • Indigenous street conditions have significant bearing on the efficacy of cycling networks.
    Portland’s residential street grid mostly consists of narrow streets without stop signs, with only occasional arterial intersections. These natural bikeways provide hundreds of miles of connections between residences and destinations, whether or not they are signed as bike routes. I believe that differential cycling mode share between cities is likely to relate to the presence of natural bikeway mileage, which is currently measured poorly, if at all.


Bike Lab’s work is supported by the Judith Lee Stronach Travel Award, College of Environmental Design, UC Berkeley.