2021

Rongxiang Su (rongxiangsu@ucsb.edu), University of California, Santa Barbara
Elizabeth McBride, University of California, Santa Barbara
Konstadinos Goulias, University of California, Santa Barbara

This paper demonstrates the use of motif and sequence analysis in tandem to analyze differences and commonalities between telecommuters and usual commuters. In terms of substantive findings, telecommuters are by far more diverse in their allocation of time to places, activities, and travel. Approximately 20% of telecommuters stay at home all day during a workday, while only 7.97% of commuters do. Telecommuters that have at least one trip during their workday accrue more vehicle miles travelled and number of trips than their commuter counterparts. However, they travel less driving alone and tend to have more complex schedules visiting more locations. Within telecommuters and commuters, however, we have substantial variation in activity participation and travel captured by the combination of motifs and sequence analysis. As expected, a substantial proportion of commuters display a morning and afternoon peaks of arriving at and departing from work, and telecommuters do not show this pattern. In addition, telecommuters do not only perform work tasks from home. Instead, during a day a high percentage travel to a variety of locations to either visit customers and/or use their spatio-temporal schedule flexibility to perform work tasks from locations other than home. In contrast, more than 80% of commuters perform work at their workplace. In addition, a slightly higher proportion of telecommuters function as the designated driver escorting other people to their activity locations.

Carole Voulgaris (cvoulgaris@gsd.harvard.edu), Harvard University
Sarah Littlefield, Harvard University

For decades, transportation planners have grappled with the potential of telecommuting and remote work to drastically alter the demand for travel. Predicted reductions in travel associated with telework technology have generally not materialized. This is partly because few workers work entirely from home. Based on our analysis of data from the American Time Use Survey, we find that it is approximately as common to split a work day between home and a workplace as it is to work entirely from home on any given work day. We estimate a logistic regression model to predict the likelihood that a worker will spend part of their work day from home and find that longer work days, the presence of a spouse or partner, and the presence of children contribute to a greater likelihood that a worker split the work day between home and a workplace. We also estimate a set of Poisson regression models to determine the relationship between working entirely or partly from home and the number of total trips a worker makes. We find that those who work entirely from home make about the same number of total daily trips as those who work entirely away from home, suggesting that those who work entirely from home tend to substitute forgone commute trips with non-work-related trips. Those who work partly from home make more total trips than either those who work entirely from home or those who work entirely away from home.

Shihang Zhang, Southwest Jiaotong University
Bin Shuai, Southwest Jiaotong University
Jie Gao, Nagoya University

As a traffic demand management policy, telework has the potential to alleviate congestion and air pollution. Because of the COVID-2019, it receives lots of attention recently. However, the substitution-complementary disputation between telework and transportation has never been settled. This study investigates the relationships between telework and commuting, and telework and work-related travel, respectively, based on the 2012-2015 German panel survey data MOP. The outcome variables include two sets (1) one-way commuting distance/duration, weekly commuting days/trips/distance/duration; (2) one-way non-commuting work-related travel distance/duration, weekly work-related days/trips/distance/duration. Both Cluster robust regression for the pooled sample and random effect regression for panel sample have been applied to model commuting and work-related travel. Overall, 19.46% of participants are teleworkers, they have longer one-way commuting duration and distance than non-teleworkers. The random effect regressions indicate telework could reduce the weekly commuting days and trips. Furthermore, it also shows a significantly negative association with weekly commuting duration, implicating a substitution effect of telework on commuting travel. On the other hand, the results show that telework is positively associated with all work-related travel, indicative of the potential of a rebound effect. The results indicate that telework has different effects on commuting and work-related travel. Meanwhile, the COVID-19 might not only stimulate more individuals teleworking, but also change the impact of telework on work-related travel. Future research should pay more attention to work-related travel.

Evelyn Blumenberg (eblumenb@ucla.edu), University of California, Los Angeles
Julene Paul, University of California, Los Angeles
Gregory Pierce, University of California, Los Angeles

Despite their negative externalities, cars provide many benefits. Chief among these is the ability to travel to destinations within a reasonable time budget. Consequently, in the U.S. most households—even low-income households—own automobiles. But technological innovations may alter this dynamic. New technology-based services and activities may reduce the advantages of private vehicle ownership, allowing households to live car-free or downsize their household vehicle fleets. In this study we investigate the relationship between these innovations and vehicle ownership using data from the 2017 National Household Travel Survey. We focus on the effects of these new services and activities on the likelihood of being a zero-vehicle household and on the relationship between household vehicles and adults. Our models indicate a positive relationship between the use of ridehail and carshare services and the likelihood of being a zero-vehicle household. The data also show a positive relationship between online shopping and working from home and the likelihood of a household having fewer vehicles than adults. Combined, the findings suggest that new technology-based activities may allow some households to eliminate or reduce their household vehicle fleets. For other households, new technology-based services may not directly affect their automobile ownership, but rather increase their access to opportunities while easing the financial burden of owning a car. Agencies and organizations should explore opportunities to better connect households—particularly households with travel and financial constraints—to technology-based services and activities.

Huyen Le (le.253@osu.edu), Ohio State University
Andre Carrel, Ohio State University
Harsh Shah, Ohio State University

E-commerce has been on the rise in recent decades, leading to substantial changes in the way people travel and perform everyday activities. The interactions between online shopping and travel behaviour have been conceptualized and quantified as four possible effects: online shopping may complement physical travel, substitutes for travel, modify travel, or have no effects (i.e., a neutral effect). This paper provides a systematic review of the literature on online shopping and travel published from 2005 to 2020. We aim to assess the amount of evidence for all four types of effects on a number of travel outcomes including trip frequency, vehicle kilometres/miles travelled, trip chaining, mode choice, and travel time. Our study provides an agenda for future research on this topic and policy implications for practitioners to incorporate e-commerce in planning for sustainable urban systems.

Rui Colaco, Universidade de Lisboa
João de Abreu e Silva, Instituto Superior Técnico Universidade de Lisboa

The permanent growth of online shopping has led to an impact on personal travel and on freight transport that is yet to be fully grasped. Previous research on the subject offers mixed findings, with several studies pointing to complementarity between online and in-store shopping, while others suggest substitution, modification, or neutrality. Using data from a 7-day shopping survey in Lisbon, Portugal, which involved 400 respondents, this paper applies structural equation modelling to explore the relationships among online shopping and in-store shopping preferences, while also considering the period of the week in which the purchases took place, since it is expected that the interaction between shopping and other personal travel behaviour varies between weekdays and weekends and therefore influences online and in-store shopping. The result shows that an online shopping preference leads to more online purchases, while in-store shopping preference leads to more in-store purchases. Furthermore, online shopping on weekdays has a positive association with both online and in-store shopping on weekends, which supports a complementarity effect. This effect is linked to a younger population, which commutes by car, and lives in less central areas. Since deliveries are becoming increasingly faster, this suggests that, if the access to the centre becomes more difficult on weekends, complementarity might give way to substitution.

Lu Xu (lux15@uci.edu), University of California, Irvine
Jean-Daniel Saphores, Institution of Transportation Studies

E-commerce has been expanding rapidly in the last decade, particularly during the Covid-19 pandemic period. Does online shopping reduce in-store shopping related travel? To answer this question, we analyze data at the household level from the 2017 National Household Travel Survey (NHTS) of the United States to understand the impact on household travel of different levels of online shopping. To reduce potential self-selection bias due to differences of household characteristics, we apply the propensity score matching (PSM) method. We find that active online shopping households engage in two more shopping activities per month on weekdays and 0.7 more shopping activities per month on weekends compared to those with less online purchases, which introduces additional 15~20 and 14 vehicle miles per month on weekdays and weekends, respectively. We also discuss the impact of online shopping on household activities like buying meals, exercise, and health care visits. We believe that our analysis results and the applied PSM method could benefit future research studies for the effects of fast-growing e-commerce on household travel, and help policy makers ultimately.

Yago Losa Rovira, Imperial College London
Ahmadreza Faghih Imani (s.faghih-imani@imperial.ac.uk), Imperial College London
Aruna Sivakumar, Imperial College London
Jacek Pawlak, Imperial College London

Virtual activity participation has spread beyond the narrow, highly skilled groups of society. Such virtual activities, including tele-working, online shopping, and virtual social interactions, are observed to explicitly impact travel behavior and activity scheduling. And yet, activity-based travel demand models do not accommodate the trade-offs between activity types and location as modulated by virtual activity participation. This paper presents a model of activity generation that captures the relationship between the three dimensions of: activity type (such as work, study, shopping), activity location (in-home, out-of-home), and activity modality (in-person, virtual). A multiple-discrete continuous extreme value (MDCEV) model structure is applied, and the empirical analysis is undertaken using the 2015 United Kingdom national household Time Use dairy Survey (UKTUS). The model results provide insights for a better understanding of the trade-offs made by individuals as they participate in and allocate time across a set of activity type-location-modality alternatives, and the heterogeneity in the trade-offs. This is a big step towards developing an activity generation model that can support the testing of a variety of scenarios related to virtual activity participation.

Tanjeeb Ahmed, University of California, Irvine
Michael Hyland (hylandm@uci.edu), University of California, Irvine

This study aims to model and analyze the use of ride-hail in trip chains considering the effects of trip chain complexity and technology usage, in addition to traditional socio-demographic, travel, and built environment characteristics. To meet this objective and investigate multiple causal relationships between trip chain complexity, technology usage, and ride-hailing usage in trip chains, the study employs structural equation modeling (SEM) techniques that incorporate latent constructs for trip chain complexity and technology use using data from the 2017 National Household Travel Survey. In addition to analyzing the effects on ride-hail, the SEM includes a second outcome variable for transit usage for comparison purposes. The results indicate significant effects of trip chain complexity on the use of transit in trip chains. Technology usage has significant direct effects on the use of ride-hail and transit. Moreover, technology usage indirectly impacts transit through trip chain complexity and the proportion of activity types (i.e. maintenance, discretionary and subsistence activity types) in a trip chain. The modeling framework and parameter estimation results in this study provide a more holistic view of the interrelationships between trip-activity chain complexity, ride-hail usage, and technology usage than is available in the existing literature. The framework and results should provide behavioral insights that have value to transportation modelers, planners, and policymakers in addition to transportation network companies.

Mootaz M.Jaff, International Islamic University Malaysia
Abdul Azeez Kadar Hamsa, International Islamic University Malaysia

Telecommuting is well suited for female employees due to their increased responsibilities in balancing both office and family matters. This paper is an attempt to evaluate the ability of female employees to telecommute in the selected business organizations in Kuala Lumpur, Malaysia. Four major job categories namely “clerical workers”, “professionals”, “associate professionals”, and “managers” were selected to ascertain how the employees under these employment categories perceive telecommuting as part of their ability to telecommute by using a questionnaire survey. A binary and an ordinal logistic regression model were used to evaluate the effects of the explanatory variables on the ability to telecommute by controlling and uncontrolling the “employing industries” and “employment categories”. The results of the model reveal that ‘prior experience working from home and ‘increased frequency of telephone usage’ almost doubles the ability to telecommute, whereas ‘time spent working with others’ and the ‘usage of the photocopier’ reduces it. The respondents employed in the ‘financial intermediation’ and ‘real estate industries’ were approximately 2.5 times more likely to be able to telecommute relative to those employed in the ‘telecommunication industry’. Managers, Associate Professionals, Professionals were 3.5, 2.3 and 1.5 times respectively more likely to be able to telecommute more frequently than Clerical workers.

Shubham Agrawal (agrawa3@clemson.edu), Clemson University
Srinivas Peeta, Georgia Institute of Technology (Georgia Tech)

This paper presents a hybrid route choice model that incorporates real-time information-induced latent cognitive and psychological effects using drivers’ physiological indicators. Electroencephalogram (EEG) data was collected for 95 participants in driving simulator experiments designed to elicit realistic route choices using network-level roadmap and responsive ambient traffic. Averaged EEG band powers in multiple brain regions were used to extract two latent cognitive/psychological effects capturing driver’s cognitive effort during and immediately after the information provision, and cognitive inattention before implementing the route choice decision. The effects of external factors (i.e., route and information characteristics, driver attributes, and situational factors) on latent variables and their combined impacts on route choice decisions were analyzed using structural equation modeling framework. Study results illustrate statistically significant impacts of driver attributes and information characteristics on latent cognitive effort and route characteristics on latent cognitive inattention. The results suggest that drivers who are more attentive to the road environment and/or exert more cognitive effort are more likely to switch from their current route. Significant direct effects of route and information characteristics but no significant direct effects of driver attributes were also observed on the route choice behavior. The study insights can aid traffic operators and information providers to incorporate human factors and psychological aspects while devising strategies for designing and disseminating real-time travel information to influence drivers’ route choices.