Measuring Changes in Multimodal Travel Behavior Resulting from Transport Supply Improvement

Despite the desired transition toward sustainable and multimodal mobility, few tools have been developed either to quantify mode use diversity or to assess the effects of transportation system enhancements on multimodal travel behaviors. This paper attempts to fill this gap by proposing a methodology to appraise the causal impact of transport supply improvement on the evolution of multimodality levels between 2013 and 2018 in Montreal (Quebec, Canada). First, the participants of two household travel surveys were clustered into types of people (PeTys) to overcome the cross-sectional nature of the data. This allowed changes in travel behavior per type over a five-year period to be evaluated. A variant of the Dalton index was then applied on a series of aggregated (weighted) intensities of use of several modes to measure multimodality. Various sensitivity analyses were carried out to determine the parameters of this indicator (sensitivity to the least used modes, intensity metric, and mode independency). Finally, a difference-in-differences causal inference approach was explored to model the influence of the improvement of three alternative transport services (transit, bikesharing, and station-based carsharing) on the evolution of modal variability by type of people. The results revealed that, after controlling for different socio-demographic and spatial attributes, increasing transport supply had a significant and positive impact on multimodality. This outcome is therefore good news for the mobility of the future as alternative modes of transport emerge.

Who’s on Board? Examining the Changing Characteristics of Transit Riders using Latent Profile Analysis

Subsidies of public transit have more than doubled since the late 1980s, with a disproportionate share of funds going to rail services. These investments have important implications, including how they affect both the composition of transit users and their travel behavior. To investigate how transit users and use are changing, we use Latent Profile Analysis and data from the 2009 and 2017 National Household Travel Surveys to examine changes in transit users in the U.S. and in five major metropolitan areas. Nationwide, we find that the share of Transit Dependents grew by 17% to account for two-thirds of all transit users in 2017. These least advantaged riders were more likely over time to reside in very poor households and to be carless. There was a corresponding decline in Occasional Transit Users, for whom transit is part of a multi-modal travel profile. Higher-income, mostly car-owning Choice Transit Riders increased slightly over time but accounted for less than one in ten transit riders in 2017. Their growth was concentrated in a few large metropolitan areas where densities and land use are most transit-supportive. While increased rail transit service has shifted riders away from buses, transit’s role as a redistributive social service that provides mobility to disadvantaged travelers has grown over time. Efforts to draw more multi-modal and car-owning travelers onto transit have been less successful. As transit systems struggle to recover riders following the pandemic, transit’s waxing role of providing mobility for those without will likely become even more prominent.

Understanding Google Location History as a Tool for Travel Diary Data Acquisition

Understanding human mobility within urban settings is fundamental for urban and transport planning. Travel demand modeling and planning typically rely on data that are collected from large-scale household travel surveys (i.e., origin–destination surveys) and compiled into single- or multiple-day travel diaries. The laborious task of collecting these data has left traditional methods with numerous limitations, resulting in significant trade-offs in regard to accuracy, sample size, and study duration, while also being vulnerable to reporting and transcription error. Rising mobile phone ownership has provided opportunities to acquire expansive cellular network data from service providers and location-based service data through smartphone applications. At the same time, the Google Maps smartphone application provides built-in infrastructure that can passively collect detailed location information from user smartphone devices. The resulting data are known as Google location history (GLH). To better understand the potential of these data offerings in transportation modeling and planning, GLH data passively collected from five different smartphones following prescribed itineraries over 12 days was evaluated. As 51% of 934 locations and 32% of 888 trips were matched to the pre-determined travel diary data, it was determined that GLH data does not currently appear to be an adequate tool for travel diary data collection. On average, locations that were missed by GLH were shorter (mean of 355 s), whereas locations that were identified were longer (mean of 762 s).

The Mismeasurement of Mobility for Walkable Neighborhoods

The major US household travel surveys do not ask the right questions to understand mobility in Walkable Neighborhoods. Yet few subjects can be more important for sustainability and real economic growth based on all things of value, including sustainability, affordability, and quality of life. Walkable Neighborhoods are a system of land use, transportation, and transportation pricing. They are areas with attractive walking distances of residential and local business land uses of sufficient density to support enough business and transit, with mobility comparable to suburbia and without owning an auto. Mobility is defined as the travel time typically spent to reach destinations outside the home, not trips among other destinations that are not related to the home base. A home round trip returns home the same day, a way of defining routine trips based on the home location. Trip times and purposes, taken together, constitute travel time budgets and add up to total travel time in the course of a day. Furthermore, for Walkable Neighborhoods, the analysis focuses on the trips most important for daily mobility. Mismeasurement consists of including trips that are not real trips to destinations outside the home, totaling 48 percent of trips. It includes purposes that are not short trips functional for walk times and mixing of different trips into single purposes, resulting in even less useful data. The surveys do not separate home round trips from other major trip types such as work round trips and overnight trips. The major household surveys collect vast amounts of information without insight into the data needed for neighborhood sustainability. The methodology of statistics gets in the way of using statistics for the deeper insights we need. Household travel surveys need to be reframed to provide the information needed to understand and improve Walkable Neighborhoods. This research makes progress on the issue, but mismeasurement prevents a better understanding of the issue.

Inferring proxy response in household travel surveys with unknown completer using a group-based choice model

A proxy response is often accepted for household travel surveys to reduce the survey cost and increase the sample size, but proxy-response biases may be introduced into the sample data. To investigate and correct the bias, completer information for the survey is important, but such information is not always available in practice. This study proposes a novel model that can be applicable in situations where completer information is unavailable. The method introduces group-decision modeling in analyzing the response choices of the household travel survey, where the survey response is considered to be a task allocation among household members. The proposed model can infer the probability of proxy response and the proxy-response bias of trip-related records without completer information. The potential of the proposed model was confirmed by application to a household travel survey in Japan. The inferred probability of the proxy response and the inferred bias without completer data demonstrated surprisingly similar results to the existing study with actual proxy-response data. Specifically, the model inferred a high probability of proxy response in young adults and a low proxy probability in middle-aged females, and the model inferred the proxy-response bias that female proxy respondents in the middle-aged group report lower trip rates than self-respondents. This method will be valuable not only in travel surveys, but also in the general research and practice of social surveys.

Application of Machine Learning to Two Large-Sample Household Travel Surveys: A Characterization of Travel Modes

Even in a context of rapidly evolving transportation and information technologies, household travel surveys remain an essential source of information for transportation planning. Moreover, as planning authorities become increasingly concerned with reducing the use of the private car, travelers’ mode choice patterns should be reexamined. In this study, a machine learning algorithm (Random Forest) was employed to characterize the use of eight different travel modes observed in two consecutive household travel surveys undertaken in Montreal, Canada. The analysis incorporated roughly 160,000 observed trips. The Random Forest algorithm was trained on the 2008 survey data and applied to the 2013 survey. The usefulness of the algorithm was evaluated using two numerical representations: the confusion matrix and the importance matrix. The results of this evaluation showed that the Random Forest algorithm could generate a detailed and precise characterization of travel submarkets for four of the most commonly observed modes of travel (auto-drive, public transit, school bus, and walk) using 11 attributes of households, persons, and trips. However, the auto-passenger mode was difficult to characterize because of its dependence on unobserved intra-household interactions. The algorithm also had difficulty identifying users of rarely observed modes (park-and-ride, kiss-and-ride, bicycle), but performed better in this regard than a traditional mode choice model. Finally, traveler’s age and the spatial orientation of origin–destination pairs were found to be decisive factors in the use of the auto-drive mode. This finding, combined with the stability of mode choice patterns observed over 5 years, highlights the difficulty of significantly reducing automobile use.