Joseph Broach

Bicycle boulevards and changes in physical activity and active transportation: Findings from a natural experiment

OBJECTIVE This study evaluates changes in physical activity and active transportation associated with installation of new bicycle boulevards. METHODS This natural experiment study uses data from a longitudinal panel of adults with children (n=353) in Portland, OR. Activity and active transportation outcomes were measured with GPS and accelerometers worn for up to 5 days in 2010-11 and 2012-13. The effect of the treatment was estimated using difference in differences estimation and multivariate regression models. RESULTS In five of the seven models, the interaction term was not significant, indicating that after controlling for the main effects of time and exposure separately, there was no correlation between being in a treatment area and minutes of moderate and vigorous physical activity (MVPA) per day, bicycling >10 min, walking >20 min, minutes of walking (if >20), or making a bike trip. Significant covariates included rain, being female, living closer to downtown, and attitudes towards bicycling, walking, and car safety. CONCLUSION This study could not confirm an increase in physical activity or active transportation among adults with children living near newly installed bicycle boulevards. Additional pre/post studies are encouraged, as well as research on the length of time after installation that behavior change is likely to occur.

Calibrated Labeling Method for Generating Bicyclist Route Choice Sets Incorporating Unbiased Attribute Variation

Discrete choice model estimation requires specification of the alternatives considered for each observed choice. In route choice problems based on real-world travel observations, generally only the chosen route is observed, and the rest of the choice set remains hidden from the analyst. In dense travel networks, thousands of paths may connect a given origin or destination, necessitating methods for generating a reasonable subset of options. A new method is proposed for generating deterministic route choice sets. The technique modifies the labeled routes method, in which multiple criteria are optimized individually to generate attractive routes. The proposed method offers two potential improvements: (a) multiple routes are generated for each label by allowing a sensitivity parameter to vary and (b) a calibration step fits the alternative shortest path deviation distribution to observed behavior. The resulting process is more flexible than traditional labeled routes, yet it maintains strong links to behavior and reduces potential attribute bias. The proposed calibrated labeling method is applied to bicyclist route choice in a dense urban network. Results suggest that the proposed technique outperforms existing methods on several key criteria. In addition, explicitly linking choice set generation to observed travel patterns creates a more intuitive behavioral link than existing strategies. The proposed method should be immediately useful for route choice modeling in similar contexts. Furthermore, the basic framework could be more broadly applicable for route choice set generation.

Travel to Common Destinations: An Exploration Using Multiday GPS Data

Decisions involving common travel patterns over time, such as traveling to and from the same destination multiple times per week or month, are likely different from decisions involving unique circumstances. Understanding of travel decisions involving common destinations could help with the development of programs, services, and infrastructure aimed at changing travel behavior. This paper explores how common travel destinations can be defined through the use of stated and revealed data (a 5-day GPS trip log). Revealed common destinations were identified through hierarchical agglomerative clustering. A 200-m maximum threshold for defining clusters was found to be appropriate in this context. Three ways of defining “common” were explored through the use of the revealed GPS data on the basis of the number of trips that ended at the cluster or the number of days that trips ended at the cluster. The analysis of travel to and from the revealed common destinations identified some potential for variability that depended on the definition. Common destinations defined by stated data (participant survey data collected before the GPS data were collected) were linked to 46% of all nonhome trip ends. The availability of these stated common destinations will reduce the need to collect information on trip purpose through the use of follow-up surveys or through the imputation of trip purpose with land use data when GPS travel surveys are processed. The analysis also provided insight into the usefulness of multiday data, thus confirming other research.

Using Predicted Bicyclist and Pedestrian Route Choice to Enhance Mode Choice Models

Recent advances in bicyclist and pedestrian route choice modeling have shown that a variety of attributes affect the paths chosen for cycling and walking. They also allow one to estimate the effect on route choice of specific network changes, such as new bicycle facilities or pedestrian crossings. Route choices do not, however, tell one anything explicit about changes in decisions to walk or cycle in the first place. Cyclists might go out of their way to use a bike lane or to avoid a busy street, but how do those same features along a potential route influence the choice to cycle instead of using another travel mode? Route choice models are applied to predict the cycling and walking routes considered for a given trip, and the resulting route-level attributes are used to predict trip mode choice. In general, existing route preferences do carry over to mode choice, but with important differences, especially for bicycle facility types and female cyclists. The results show that available off-street paths and low-traffic on-street routes not only draw cyclists from other facilities but also make prospective riders more likely to cycle on a given trip. Gender differences are found for decisions to bicycle, with women showing a lower propensity than men to cycle on a similar trip and also stronger sensitivity to the availability of routes with lower traffic stress. Traffic-calmed streets, such as bicycle boulevards, may be particularly important in reducing the observed bicycling gender gap for everyday travel.

Analysis of Short-Duration Unscheduled Absences of Transit Operators: TriMet Case Study

Factors that contribute to short-duration (1 to 3 days) unscheduled absences of operators at the Tri-County Metropolitan Transportation District of Oregon (TriMet), the transit provider for the Portland, Oregon, metropolitan region, are analyzed. The analysis draws on a wide array of operator-specific information recovered by the technologies of transit intelligent transportation systems in combination with information from the agencys human resources, scheduling, incident, and customer relations databases. The likelihood of an absence is estimated in relation to personal characteristics, employment status, aspects of assigned work, service delivery, performance indicators, temporal factors, and customer feedback. The findings can be used directly to support extraboard planning practices. More generally, the findings point to changes in policies and practices that could reduce the incidence of short-duration absences.

Existence and Use of Low-Pollution Route Options for Observed Bicycling Trips

Do routes with lower doses of air pollution exist in real-world bicycling networks, and do bicyclists actually use those routes? Low-pollution-dose alternative routes for a sample of urban cycling trips were modeled and compared with shortest paths. Bicyclists’ actual route choices on the same trips were observed with the use of GPS data and compared with the low-dose and shortest paths alternatives. With use of past studies of pollution exposure levels and simplified ventilation rates, link-inhaled doses of air pollution were estimated. Findings suggest that a majority of trips have lower-dose alternatives to the shortest path, with a 12% average dose reduction. Cyclists tend to choose routes with pollution concentrations between those of shortest paths and minimum-dose routes, but they also travel considerably farther, leading to total inhaled doses that are higher than on either alternative route. People’s seeming avoidance of nontraffic factors such as hills, excess turns, and difficult intersections leads to longer than optimal detours from a pollution avoidance perspective. Bike paths and bike boulevards (traffic-calmed streets with bicycle priority), as well as denser street grids, appear to provide effective low-pollution alternatives, although such routes tend to encourage excess detours that can add to total inhaled dose. Bike lanes can draw cyclists onto more polluted routes in some circumstances, with poor pollution inhalation outcomes. Overall, excess doses did seem to be a common problem for this sample of cyclists on a real-world network. The study’s findings support policies that provide dense networks of attractive facilities that encourage cyclists to choose direct, lower-pollution routes.

Current Efforts to Make Bikeshare More Equitable

The number of public bikeshare systems has been increasing rapidly across the United States over the past 5 to 10 years. To date, most academic research around bikeshare systems in the United States has focused on the logistics of planning and operationalizing successful systems. Investigations of system users and effects on the local community are less common, and studies that are focused on efforts to engage underserved communities in bikeshare systems are rarer still. This paper relies on a survey of representatives from 56 U.S. bikeshare systems to better understand and document current approaches toward serving low-income and minority populations. The survey asked about equity policies and metrics, the degree to which equity considerations affect a variety of system practices, what the existing barriers to utilizing bikeshare are for target populations, and what challenges the bikeshare system entity faces in addressing those barriers. Results indicate that nearly one in four bikeshare systems has written policies around equity; nearly half of bikeshare systems with more than 500 bikes have such policies. However, many more systems consider equity in various aspects of their systems. Equity considerations affected station siting, fee structure and payment systems, and promotion and marketing in a majority of systems (68%, 72%, and 57%, respectively), and operations and data collection and analysis, though to a lesser extent (42% each). Bikeshare systems reported cost, access, and outreach as the largest barriers to equity, in addition to overall funding and staff levels.