National Household Travel Survey Research Articles

Inequality in activity participation: Multidimensional disadvantages and daily trips by trip purpose and trip day

This paper investigated how daily trips are associated with multidimensional disadvantages in demographic characteristics, socioeconomic status, transportation barriers, and internet use based on the 2017 U.S. National Household Travel Survey. We examined how these disadvantages affected weekday and weekend trips for work, recreation, and social participation purposes. Accounting for transit services and walkability at the metropolitan area level, the results of multilevel ordered and binomial logistic regression revealed that although the disadvantages were negatively associated with the likelihood of making any trips, their associations varied by the type of disadvantages, trip purpose, and trip day. Having a lower income was associated with a higher likelihood of making a weekday social participation trip and a weekend work trip. No internet use was more consistently related to a lower likelihood of making any weekday trips. A lack of car ownership was negatively associated with social participation trips, and the perceived financial burden of trips was negatively related to recreational trips. We discuss policy implications for reducing the effects of these disadvantages on trip likelihood.

Electric vehicle market potential and associated energy and emissions reduction benefits

In this paper, a methodological framework is proposed to assess the potential of electric vehicle (EV) penetration and corresponding reduction in energy use and emissions in Georgia, U.S., using 2017 National Household Travel Survey data. A two-phase, data-driven model assesses the potential for household purchases of EVs and the assignment of EVs to household trips. Households sharing the highest similarities are selected as candidates for EV purchases, with household trips identified as EV-amenable or not. Potential EV-purchasing families were also matched to specific EV makes and models. Energy use, greenhouse gas emissions, and criteria air pollutants were analyzed and compared for all original trips and for those trips that shifted to EVs. By comparing against two traditional averaging methods, the framework demonstrates an advancement that helps to avoid the overestimation of EV benefits. By integrating household-level demographics and trip-level attributes from open-source travel survey data in EV adoption and trip assignment, this paper demonstrates the benefits associated with EV adoption (a potential 45% reduction in energy consumption and 30% reduction in greenhouse gas emissions), moreover, the proposed methodology could serve as an innovative framework that is scalable and transferable to predict the future market penetration and actual on-road EV activities under various contexts.

Health impacts of nonmotorized travel behavior and the built environment: Evidence from the 2017 National Household Travel Survey

IntroductionThe health impacts of nonmotorized travel behavior (i.e., walking and bicycling) and the built environment have been widely investigated in the past. However, nationwide studies probing the association between the neighborhood built environment and individuals' nonmotorized travel and health status remain scarce. Utilizing national-level databases from the U.S., the present study examines the role of nonmotorized travel and neighborhood-level built environment attributes in residents’ self-assessed overall health status. MethodsThis research relies on data from the 2017 National Household Travel Survey and employs ordered probit modeling techniques to relate self-assessed overall health status to various variables representing the extent of nonmotorized travel and physical activity as well as the built and social environment attributes of the residential location. ResultsThe results indicate that individuals’ self-assessed health status is associated with i) their person-level attributes including the levels of their walking and bicycling travel and other physical activity; ii) the social environment attributes of both their household and their residential neighborhood including income and vehicle ownership levels; and iii) the built environment attributes of their neighborhood including the extent of: compactness, mixed-use development, transit accessibility, and automobile-friendliness of the road network within the neighborhood. Furthermore, average marginal effects computations indicate that the most important neighborhood built environment factor in increasing the likelihood of reporting a very good or excellent health status is higher levels of transit accessibility. ConclusionsThe findings of this study suggest that more effective policies to promote individuals' health through changes targeting their travel behavior and their neighborhood's built environment may be the ones that: a) encourage walking and bicycling travel, and b) foster neighborhood designs that support more active lifestyles through increased levels of transit accessibility, activity density, and land use diversity, as well as through decreased levels of automobile-friendliness of the street network.

Disparities in Activity and Traffic Fatalities by Race/Ethnicity

Traffic fatalities remain a major public health challenge despite progress made during recent decades. This study develops exposure-based estimates of fatalities per mile traveled for pedestrians, cyclists, and light-duty vehicle occupants and describes disparities by race/ethnicity, including a subanalysis of fatality rates during darkness and in urban areas. Estimates of person-miles traveled by mode and race/ethnicity group were derived from the 2017 National Household Travel Survey using replicate weights. Three-year average (2016‒2018) traffic fatalities were measured by mode and race/ethnicity group with the U.S. Fatality Analysis Reporting System. Fatality rates per mile traveled and CIs were calculated for each subgroup as well as separately for trips occurring during darkness and in urban areas. Analysis was conducted in 2021‒2022. Exposure to traffic fatality differs by race/ethnicity group and by mode, indicating that adjustment for differential exposure is needed when estimating disparities. The authors find that fatality rates per 100 million miles traveled are systematically higher for Black and Hispanic Americans for all modes and notably higher for vulnerable modes (e.g., Black Americans died at more than 4 times the rate for White Americans while cycling, 33.71 [95% CI: 21.84, 73.83] compared with 7.53 [95% CI: 6.64, 8.69], and more than 2 times the rate while walking, 40.92 [95% CI: 36.58, 46.44] compared with 18.77 [95% CI: 17.30, 20.51]). Previous estimates that do not adjust for differential exposure may underestimate disparities by race/ethnicity. Observed disparities remained when considering only urban areas and appear to be exacerbated during darkness. Traffic fatalities are a substantial and preventable public health challenge in America. Black and Hispanic Americans have higher traffic fatality rates per mile traveled than White Americans across the transportation system, requiring urgent attention.

Time on the road and the price of Gasoline: Evidence from ATUS and NHTS

What happens when motor fuel prices rise? In the US, the National Household Travel Survey (NHTS) collects information about car ownership and use, and travel during a typical day, for over 100,000 households. It is conducted only once every eight years, and does not include a longitudinal component, making it difficult to observe drivers’ adjustments to changing gasoline prices. We experiment with combining the 2017 NHTS with eight waves of the American Time Use Survey (ATUS), which tracks trips, time spent traveling, and other characteristics of each trip. We find that the two datasets document remarkably similar behaviors—whether we use the eight waves of the ATUS or limit the comparison to the period when the two surveys overlap (April 2016 to April 2017). They also document similar responsiveness to prices—at least for the decision to take a car or a public transit trip. By contrast, minutes on the road and miles in the NHTS appear to be strongly responsive to gasoline prices, whereas minutes on the road from the ATUS are unrelated to fuel prices, despite the much greater price variation therein. The results are robust to extensive checks and efforts to reduce measurement error in gasoline prices. A 25-cent increase in the price of gasoline is predicted to reduce CO2 emissions by 1–5%.

Electric vehicles in multi-vehicle households

ABSTRACT This paper uses U.S. nationally representative data from the 2017 National Household Travel Survey to present a series of facts about electric vehicles (EVs) in multi-vehicle households. First, as of the time of the survey, 89% of households with an EV also had a non-electric vehicle in addition to the EV. Second, 60% of households with an EV also had a non-electric SUV, truck, or minivan. Third, 66% of households with an EV also had a non-electric vehicle that was driven more miles per year. The paper argues that these patterns have significant implications for the environmental impact of EVs and underscore the importance of better understanding how multi-vehicle households substitute between vehicles.

A structural analysis of the work tour behavior of transit commuters

Our knowledge of complex travel behavior associated with transit commuting is limited. While chaining activities into tours has been a frequent behavior of travelers in general, and in some cases of work commuters, it is unclear how successful public transit has been in providing the accessibility that is needed to link multiple activities, especially on work tours. To address this knowledge gap, a structural model was developed for transit commuters, which allows for the characterization of commuters based on the complexity of work tours and an assessment of the influence of household and person-level socio-demographic characteristics, built environment variables, and activity-travel demand on tour complexity. Using data from the 2017 National Household Travel Survey, this study’s results suggested that married men with no children and high vehicle ownership living in low-density areas tend to make simple work tours while single women with children who live in high-density neighborhoods were more likely to make complex work tours. Also, white men with higher incomes and higher education living in denser areas were more likely to make complex tours with work-based sub-tours. Denser residential neighborhoods, flexible work schedules, and vehicle availability on work tours were observed to increase the propensity of making complex tours. The findings of this study can assist transit agencies or planning organizations in identifying transit commuters who have complex travel needs or whose circumstances hold potential to yield greater benefits from transit usage in work tours and thus to formulate policies directed at better work and non-work travel and activity linkages.

Robust Bayesian inference for Big Data: Combining sensor-based records with traditional survey data

Big Data often presents as massive nonprobability samples. Not only is the selection mechanism often unknown but larger data volume amplifies the relative contribution of selection bias to total error. Existing bias adjustment approaches assume that the conditional mean structures have been correctly specified for the selection indicator or key substantive measures. In the presence of a reference probability sample, these methods rely on a pseudolikelihood method to account for the sampling weights of the reference sample, which is parametric in nature. Under a Bayesian framework, handling the sampling weights is an even bigger hurdle. To further protect against model misspecification, we expand the idea of double robustness such that more flexible nonparametric methods as well as Bayesian models can be used for prediction. In particular, we employ Bayesian additive regression trees which not only capture nonlinear associations automatically but permit direct quantification of the uncertainty of point estimates through its posterior predictive draws. We apply our method to sensor-based naturalistic driving data from the second Strategic Highway Research Program using the 2017 National Household Travel Survey as a benchmark.

Spatiotemporal analysis of activity-travel fragmentation based on spatial clustering and sequence analysis

In this study using data from the 2017 National Household Travel Survey in California from 26,078 survey participants, sequence analysis is used to estimate a fragmentation indicator of people's daily schedules. Then, spatial clustering is used to find groups of observations with similarly high or low fragmentation using the longitude and latitude of their residential locations. This is followed by a hierarchical sequence clustering within each spatial cluster to identify distinct patterns of time allocation. Using the Local Indicator of Spatial Association (LISA) we find a large portion (approximately 30%) of the sample with significant spatial clustering of fragmentation. We also find systematic and significant differences in membership to these clusters based on land use, county of residence, household and personal characteristics, and travel modes used. Sequence analysis pattern recognition within LISA spatial clusters shows systematically repeating time allocation patterns that include typical work and school schedules as well as staying at home patterns. However, each spatial LISA cluster is composed of different time allocation clusters. All this analysis taken together points out substantial and measurable heterogeneity in spatial clustering of fragmentation and the need for customized policy actions in different geographies.

Prediction of Charging Requirements for Electric Vehicles Based on Multiagent Intelligence

This study investigates the spatiotemporal distribution of electric vehicle (EV) charging demands and operating efficiency of the charging system. The travel behavior of EV drivers is analyzed by considering the heterogeneity of range anxiety and bound rationality. Building on existing charging choice models, our more holistic perspective of charging demand distribution is obtained through multiagent system (MAS) modeling. In our study, the charging demand distribution in different areas is compared by considering two charging price schemes. The performance of the charging system is then evaluated based on three indicators, the charging request rejection rate, charging pile utilization rate, and charging load deviation, thereby verifying the effectiveness of our charging demand prediction model. This is done using multiagent-based simulations applied to the National Household Travel Survey (NHTS) 2017 dataset. The results are analyzed according to different key indicators, mainly indicating (a) the limitations of the time-of-use (TOU) pricing strategy in reducing the peak-valley difference, (b) the transferability of charging demand among different functional areas, and (c) that there is less demand for charging piles in the workplace, as users mainly rely on home charging. These results facilitate further analyses to help the design and operation of the EV charging infrastructure.

To What Extent May Transit Stop Spacing Be Increased before Driving Away Riders? Referring to Evidence of the 2017 NHTS in the United States

With the emergence of ride-sourcing and ride-splitting services, more options are available to support shifts away from transit, where maintaining transit ridership increases requirements for transit service quality, so as to promote high-capacity and sustainable transport systems. In this endeavor, proper transit stop spacing is critical for both service accessibility and in-vehicle trip efficiency, as well as operation cost. This research explores acceptable stop spacing for three kinds of transit services from the perspective of travel behavior, drawing on the 2017 National Household Travel Survey in the United States. A stochastic frontier model is developed to infer passengers’ unobservable vertex of acceptable transit access times on the basis of observed walk time, which can be converted to the tolerance with respect to stop spacing with the average walking speed. Significant explanatory variables on the vertex of acceptable transit stop spacing are further identified with their quantified impacts, including household density, household income, wait time, trip distance, transfer, and maintenance purpose, while the inefficiency variance is significantly related to traveler age, wait time, secondary walk time, and trip frequency. Recommended response strategies follow. Findings from this study provide insights, guidelines, and implementation plans for different transit agencies when considering stop spacing redesign, to strengthen transit service appeal and to promote cooperative and sustainable multi-modal urban transport systems.

Transportation Patterns of Adults With Travel-Limiting Disabilities in Rural and Urban America.

IntroductionLack of transportation is a significant barrier to community participation for many disabled adults. Living in a rural area introduces additional transportation barriers, such as having to travel long distances to access services or socialize, and limited public transit options. While the importance of transportation access is clear, the mix of different transportation options used by people with disabilities to participate in their communities is less understood, particularly among those who do not or cannot drive.MethodsWe used data from the 2017 National Household Travel Survey to explore transportation behaviors among disabled adults in rural and urban areas and by four regions across the United States. We explored differences by transportation modalities (e.g., driver, passenger, public transportation, taxi/uber, walk) and trip purposes (e.g., social, independent living, healthcare, work). Our sample included 22,716 adults with travel-limiting disabilities.ResultsSeveral geographic differences emerged among non-drivers. Rural non-drivers were less likely to take any trip, particularly for social activities, and reported using less public transportation or walking/rolling than urban non-drivers. Further, respondents from the Northeast were more likely to report using public transportation and walking/rolling options, relative to the Midwest, South, and West. Overall, disabled rural adults reported lower odds of giving up driving, even after controlling for socio-demographic and health characteristics.DiscussionThese findings highlight the relative importance of different transportation modalities for participating in activities and the continued reliance upon personal vehicles, either as a driver or passenger, especially among rural disabled residents. Potential policy insights are discussed.

How Sustainable Is People’s Travel to Reach Public Transit Stations to Go to Work? A Machine Learning Approach to Reveal Complex Relationships

Several previous studies examined the variables of public-transit-related walking and privately owned vehicles (POVs) to go to work. However, most studies neglect the possible non-linear relationships between these variables and other potential variables. Using the 2017 U.S. National Household Travel Survey, we employ the Bayesian Network algorithm to evaluate the non-linear and interaction impacts of health condition attributes, work trip attributes, work attributes, and individual and household attributes on walking and privately owned vehicles to reach public transit stations to go to work in California. The authors found that the trip time to public transit stations is the most important factor in individuals’ walking decision to reach public transit stations. Additionally, it was found that this factor was mediated by population density. For the POV model, the population density was identified as the most important factor and was mediated by travel time to work. These findings suggest that encouraging individuals to walk to public transit stations to go to work in California may be accomplished by adopting planning practices that support dense urban growth and, as a result, reduce trip times to transit stations.

Nonlinear Relationships between Vehicle Ownership and Household Travel Characteristics and Built Environment Attributes in the US Using the XGBT Algorithm

In the United States, several studies have looked at the association between automobile ownership and sociodemographic factors and built environment qualities, but few have looked at household travel characteristics. Their interactions and nonlinear linkages are frequently overlooked in existing studies. Utilizing the 2017 US National Household Travel Survey, the authors employed an extreme gradient boosting tree model to evaluate the nonlinear and interaction impacts of household travel characteristics and built environment factors on vehicle ownership in three states of the United States (California, Missouri, and Kansas) that are different in population size. To develop these models, three main XGBT parameters, including the number of trees, maximal depth, and minimum rows, were optimized using a grid search technique. In California, the predictability of vehicle ownership was driven by household travel characteristics (cumulative importance: 0.62). Predictions for vehicle ownership in Missouri and Kansas were dominantly influenced by sociodemographic factors (cumulative importance: 0.53 and 0.55, respectively). In all states, the authors found that the number of drivers in a household plays a vital role in the vehicle ownership decisions of households. Regarding the built environment attributes, deficiencies in cycling infrastructure were the most prominent attribute in predicting household vehicle ownership in California. This variable, however, has threshold connections with vehicle ownership, but the magnitude of these relationships is small. The outcomes imply that improving the condition of cycling infrastructure will help reduce the number of vehicles. In addition, incentives that encourage the households’ drivers not to buy new vehicles are helpful. The outcomes of this study might aid policymakers in developing policies that encourage sustainable vehicle ownership in the United States.

Trip deprivation among older adults in the context of the capability approach

Transportation researchers have been increasingly interested in integrating the capability approach (CA) into transportation equity analysis during the past decade. The limited number of empirical studies published in that context have assessed individuals' capabilities (i.e., freedom or opportunities) and functionings (i.e., achievements) by variables such as accessibility, transit availability, ability to use transit, ideal travel options versus actual travel, etc. This paper argues that trip deprivation can legitimately reflect both functioning and capability when deprivation is carefully defined. Using data from a recent survey of 3003 older adults in New Jersey, it subsequently compares the personal, household, and locational characteristics of the trip-deprived with the characteristics of other older adults with a multi-level logit model. The trip-deprived are defined in this research as people who had forgone trips for lacking transportation at least sometimes during a six-month period. Consistent with mainstream CA literature, evidence was found that trip deprivation is significantly explained by personal attributes, household resources, and relational support, but not by location-related variables such as jobs or transit stations/stops in municipalities. The results provide empirical evidence supporting recent theoretical studies in the CA context asserting that accessibility in equity analysis should be defined as the ease of travel for individuals instead of aggregation of proximate activities. The paper concludes by suggesting that meaningful questions about potential trips, deprived trips, and the importance of visiting specific activities should be included in the National Household Travel Survey as well as various metropolitan household travel surveys for improved transportation equity analysis.

Exploring the role of ride-hailing in trip chains

Ride-hailing can potentially provide a variety of benefits to individuals who need to chain several activities together within a single trip chain, relative to other travel modes. Using household travel diary/survey data, the goal of this study is to assess the role ride-hailing currently plays within trip chains. Specifically, the study aims to determine, within trip chains, who uses ride-hailing services, for what trip/activity purposes, and to/from what types of areas, as well as the characteristics of trip chains that involve ride-hailing segments. To meet these objectives, the study estimates a binary logit model using 2017 National Household Travel Survey data, where the dependent variable denotes the inclusion of at least one ride-hailing trip within a trip chain. Similar to the non-trip-chaining ride-hailing literature, this study indicates that trip chains with ride-hailing legs are positively associated with travelers who are younger, live in high-income households, frequently use transit, and reside in high-density areas. However, this study includes novel findings indicating statistically significant relationships between ride-hailing and trip chains that end in healthcare and social/recreational activities. Moreover, trip chains with ride-hailing tend to have fewer stops and longer activity durations than trip chains without ride-hailing. This study also includes nested logit choice models, wherein the dependent variable denotes the primary mode (ride-hailing, transit, personal vehicle, or non-motorized transport) of a trip chain. These model results provide additional insights into the role of ride-hailing within trip chains, as they allow for cross-mode comparisons. The paper discusses the potential transportation planning and policy implications of the model results as well as future research directions.

Reliability-as-a-Service Usage of Electric Vehicles: Suitability Analysis for Different Types of Buildings

The use of electric vehicles (EVs) to provide different grid services is becoming possible due to the increased penetration levels, mileage efficiencies, and useable battery sizes of EVs. One such application is providing reliability-as-a-service (RaaS) during short-term power outages. Instead of using a dedicated backup power source, EVs can be contracted to provide RaaS, which is an environmentally friendly solution with benefits for both building owners and EV owners. However, the presence of EVs at a particular location during different hours of the day and the availability of energy from EVs is uncertain. Therefore, in this study, a suitability analysis is performed concerning the use of EVs to provide RaaS for different types of buildings. First, the National Household Travel Survey (NHTS) survey data are used to estimate driver behavior, such as arrival/departure times, daily mileage, and traveling duration. Then, the usable battery size and mileage efficiency of EVs is extracted from the database of commercially available EVs. Based on these parameters, the daily energy consumption and available energy of EVs to provide RaaS are estimated. A suitability analysis is conducted for residential, commercial/industrial, and mixed buildings for both weekdays and holidays. The participation ratio of EV owners is varied between 10 and 90%, and nine cases are simulated for commercial/industrial buildings and multi-unit residential buildings. Similarly, the ratio of home-based EVs is varied between 5 and 50%, and 10 cases are tested for mixed buildings. The analysis shows that mixed buildings are the most suitable, while commercial/industrial buildings are the least suitable for using EVs to provide RaaS. To this end, an index is proposed to analyze and determine the desired ratio of EVs to be contracted from homes and workplaces for mixed buildings. Finally, the impact of EV fleet size on the available energy for RaaS is also analyzed.

Classification of Imbalanced Travel Mode Choice to Work Data Using Adjustable SVM Model

The investigation of travel mode choice is an essential task in transport planning and policymaking for predicting travel demands. Typically, mode choice datasets are imbalanced and learning from such datasets is challenging. This study deals with imbalanced mode choice data by developing an algorithm (SVMAK) based on a support vector machine model and the theory of adjusting kernel scaling. The kernel function’s choice was evaluated by applying the likelihood-ratio chi-square and weighting measures. The empirical assessment was performed on the 2017 National Household Travel Survey–California dataset. The performance of the SVMAK model was compared with several other models, including neural networks, XGBoost, Bayesian Network, standard support vector machine model, and some SVM-based models that were previously developed to handle the imbalanced datasets. The SVMAK model outperformed these models, and in some cases improved the accuracy of the minority class classification. For the majority class, the accuracy improvement was substantial. This algorithm can be applied to other tasks in the transport planning domain that deal with uneven data distribution.

Modeling household online shopping demand in the U.S.: a machine learning approach and comparative investigation between 2009 and 2017.

Despite the rapid growth of online shopping and research interest in the relationship between online and in-store shopping, national-level modeling and investigation of the demand for online shopping with a prediction focus remain limited in the literature. This paper differs from prior work and leverages two recent releases of the U.S. National Household Travel Survey (NHTS) data for 2009 and 2017 to develop machine learning (ML) models, specifically gradient boosting machine (GBM), for predicting household-level online shopping purchases. The NHTS data allow for not only conducting nationwide investigation but also at the level of households, which is more appropriate than at the individual level given the connected consumption and shopping needs of members in a household. We follow a systematic procedure for model development including employing Recursive Feature Elimination algorithm to select input variables (features) in order to reduce the risk of model overfitting and increase model explainability. Among several ML models, GBM is found to yield the best prediction accuracy. Extensive post-modeling investigation is conducted in a comparative manner between 2009 and 2017, including quantifying the importance of each input variable in predicting online shopping demand, and characterizing value-dependent relationships between demand and the input variables. In doing so, two latest advances in machine learning techniques, namely Shapley value-based feature importance and Accumulated Local Effects plots, are adopted to overcome inherent drawbacks of the popular techniques in current ML modeling. The modeling and investigation are performed at the national level, with a number of findings obtained. The models developed and insights gained can be used for online shopping-related freight demand generation and may also be considered for evaluating the potential impact of relevant policies on online shopping demand.

Impact of ride-hailing usage on vehicle ownership in the United States

• Household-level data mainly from the 2017 NHTS are used, covering 42 US metro areas. • The recursive effect between ride-hailing and vehicle ownership was considered. • Regular/active ride-hailing users were negatively related to the change in ownership. • Policy implications on vehicle ownership and ride-hailing are discussed. Ride-hailing as an innovative urban mobility service has received increasing attention in the recent decade. Using data from the 2017 National Household Travel Survey, we investigate the relationship between ride-hailing usage and household vehicle ownership from 42 metropolitan areas across the United States. Previous studies have shown mixed findings on this relationship. Our study contributes to the literature by estimating a bivariate ordered probit model with considerations of the recursive effect between ride-hailing usage and household vehicle ownership. We find that compared to occasional users, regular and active users are more likely to own fewer vehicles. Also, we find no significant difference in possessing vehicles between regular and active users.