US National Household Travel Research Articles

Temporary versus Permanent Pandemic Transit Leavers: Findings from the 2022 US National Household Travel Survey

Using data from the 2022 National Household Travel Survey, I explore the socio-demographic characteristics of Americans who reduced their use of public transit during the latter stages of the COVID-19 pandemic. I also examine differences between travelers whose reduced transit use was temporary versus permanent. Using adjusted Wald tests and multinomial logistic regression, I find significant differences between people who did not leave transit and those who did, as well as between temporary and permanent transit leavers. Notably, owning a vehicle, having a disability, and working from home were associated with leaving public transit permanently rather than temporarily.

Modeling EV Charging Station Loads Considering On-Road Wireless Charging Capabilities

Electric vehicle (EV) customers are expected to charge EV batteries at a rapid EV charging station or via on-road wireless EV charging systems when possible, as per their charging needs to successfully complete any remaining trips and reach their destination. When on-road wireless EV charging systems are considered as an alternative charging method for EVs, this can affect the load of a rapid EV charging station in terms of time and magnitude. Hence, this paper presents a probabilistic framework for estimating the arrival rate of EVs at an EV rapid charging station, considering the availability of on-road wireless charging systems as an alternative charging method. The proposed model incorporates an Electric Vehicle Decision Tree that predicts the times when EVs require rapid charging based on realistic transportation data. A Monte Carlo simulation approach is used to capture uncertainties in EV user decisions regarding charging types. A queuing model is then developed to estimate the charging load for multiple EVs at the charging station, with and without the consideration of on-road EV wireless charging systems. A case study and simulation results considering a 32-bus distribution system and the US National Household Travel Survey (NHTS) data are presented and discussed to demonstrate the impact of on-road wireless EV charging on the loads of an rapid EV charging station. It is observed that having on-road wireless EV charging as complementary charging to EV charging stations helps to significantly reduce the peak load of the charging station, which improves the power system capacity and defers the need for system upgrades.

Forecast Performance of Metropolitan Trip Generation Models Statistically Updated with US National Household Travel Survey Data

Forecast Performance of Metropolitan Trip Generation Models Statistically Updated with US National Household Travel Survey Data

Targeting Sustainable Transportation Development: The Support Vector Machine and the Bayesian Optimization Algorithm for Classifying Household Vehicle Ownership

Predicting household vehicle ownership (HVO) is a crucial component of travel demand forecasting. Furthermore, reliable HVO prediction is critical for achieving sustainable transportation development objectives in an era of rapid urbanization. This research predicted the HVO using a support vector machine (SVM) model optimized using the Bayesian Optimization (BO) algorithm. BO is used to determine the optimal SVM parameter values. This hybrid model was applied to two datasets derived from the US National Household Travel Survey dataset. Thus, two optimized SVM models were developed, namely SVMBO#1 and SVMBO#2. Using the confusion matrix, accuracy, receiver operating characteristic (ROC), and area under the ROC, the outcomes of these two hybrid models were examined. Additionally, the results of hybrid SVM models were compared with those of other machine learning models. The results demonstrated that the BO algorithm enhanced the performance of the standard SVM model for predicting the HVO. The BO method determined the Gaussian kernel to be the optimal kernel function for both datasets. The performance of the SVM#1 model was improved by 4.27% and 5.16% for the training and testing phases, respectively. For SVM#2 model, the performance of this model was improved by 1.20% and 2.14% for the training and testing phases, respectively. Moreover, the BO method enhanced the AUC of the SVM models used to predict the HVO. The hybrid SVM models also outperformed other machine learning models developed in this study. The findings of this study showed that SVM models hybridized with the BO algorithm can effectively predict the HVO and can be employed in the process of travel demand forecasting.

Rideshare use among people with disabilities: Patterns and predictors based on a large nationally representative survey

People with disabilities have more difficulty getting around due to lack of transportation. Rideshare has the potential to improve mobility by addressing many limitations of fixed-route and paratransit services. However, it is unclear whether people with disabilities have access to and are benefiting from rideshare or how those benefits differ by type of disability. The purpose of this study was to examine rideshare use among people with disabilities and identify personal, household, and community characteristics associated with adoption and frequency of rideshare use. We leveraged data from the most recent US National Household Travel Survey and ran regression models to study the separate processes of adoption and frequency. We found that having a disability of any type was associated with a lower likelihood of adopting rideshare. However, among those who do adopt rideshare, having any type of disability was associated with higher rate of trip taking compared to people without disabilities. Higher frequency of rideshare use among people with disabilities was associated with being in better health, being younger, Black, better educated, employed or student, and living in densely populated areas. Our results show how rideshare behavior differs for people with disabilities compared to those without and the role of assistive devices used by people with different types of disabilities. To develop rideshare as an equitable mobility option, transportation planners, policy makers, disability advocates, and the rideshare industry need to address both the lack of access to rideshare and the socio-economic factors impacting rideshare use among people with disabilities.

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.

What is your shopping travel style? Heterogeneity in US households’ online shopping and travel

Although online shopping has been shown to interact with travel behavior in several ways, most studies to date focused on individual-level behavior in non-representative, geographically limited samples, making it difficult to derive clear profiles of shoppers. Using the 2017 US National Household Travel Survey, which included information on online shopping frequency, we estimated a latent class model to identify different shopper types (classes) that exhibited distinct travel and online shopping behavior. We found four classes: time-pressured shoppers, dual-channel shoppers, traditional shoppers, and infrequent shoppers and travelers, that differed in terms of sociodemographic characteristics and stages of life. Our results suggest that this heterogeneity may extend to the effects of online shopping on shopping travel, for instance, with substitution effects being dominant for one class whereas complementarity effects may be dominant for another. These findings can inform the design of tailored policies to mitigate the sustainability impacts of online shopping while also addressing the various classes’ needs. Further research is needed to untangle the complex relationships between online shopping and travel behavior, especially considering this heterogeneity and the modification effect.

Are cycling and walking good for all? Tracking differences in associations among active travel, socioeconomics, gentrification, and self-reported health

IntroductionWhile the health benefits of cycling and walking have been well established, questions remain about whether these benefits hold in varying socioeconomic contexts, including across demographic groups and in the context of neighborhood change. This study examines this relationship, identifying associations between cycling or walking and self-reported health, whether socioeconomic status moderates these associations, and whether gentrification influences the potential moderating effects. MethodsThis study uses the 2017 US National Household Travel Survey, subset to adults who lived in central cities (n = 91,541). Weighted logistic regression models with interaction terms were fit to estimate self-reported health status separately for cycling and walking. Gentrification was measured using an indicator based on previous research using US Census data. ResultsCycling in the previous week and each additional walking trip were associated with higher odds of reporting better health. Socioeconomic status moderated the positive associations between active transportation and health for a few variables. Cyclists of color had smaller health effects from cycling for utilitarian trips, while race and ethnicity had only a marginal moderating effect on the association between walking and health. Gentrification was an insignificant moderating factor in general. ConclusionsFindings suggest planning efforts that continue to support programs that promote cycling and walking are crucial tools in the public health toolbox. The health gains from active transportation might be experienced in a variety of neighborhood contexts. Nevertheless, infrastructure investments and policy must be attentive to inequities across neighborhoods.

Assessing the VMT effect of ridesourcing services in the US

Abstract The net effect of ridesourcing (RS) services on vehicle miles traveled (VMT) is ambiguous. With 2017 US National Household Travel Survey data, this study measures and compares the heterogeneous VMT effects of RS across population groups with various levels of household vehicle access and RS usage. A propensity score matching method was implemented to match RS non-users, occasional users, and frequent users based on observable sociodemographic traits. The results suggest that among drivers with household vehicle access, frequent RS users generate the least VMT, but occasional users actually generate more VMT. Those without a driver’s license or a household vehicle generate lower VMT while use more transit, and increasing RS use uniformly increases their overall VMT generation. We estimate that overall, RS generated a net increase of 7.8 million daily VMT in the US, compared with a counterfactual case in which all NHTS 2017 respondents were non-users of RS.

Gender Responsiveness in Public Transit: Evidence from the 2017 US National Household Travel Survey

Abstract Public transportation plays an important role in urban sustainability. To increase public transit usage, it is essential to understand the underlying reasons that discourage people from us...

The evolution, usage and trip patterns of taxis & ridesourcing services: evidence from 2001, 2009 & 2017 US National Household Travel Survey

Given the rapid adoption of ridesourcing services (RS), it is critical for transportation planners and policymakers to understand their impacts and keep policies up to date. This study contributes to the literature by using representative samples captured in the 2001, 2009 and 2017 National Household Travel Surveys to explore how taxis and ridesourcing (T/R) services have evolved and shaped people’s travel behavior pre- and post-disruption at the US national level. It characterizes and visualizes the asymmetries in demand spatially and temporally for T/R trips, showing that ridesourcing has greatly increased T/R trips from flexible and optional activity locations to home, which vary by times of day. It also characterizes tours involving T/R services, showing that while simple optional tours (such as home–recreation–home) represent the largest share of tours involving T/R, the fastest growth has been in simple mandatory tours (such as home–work–home). Tours involving T/R grew from 0.4% of all tours in 2009 to 1% of all tours in 2017, mostly within densely populated and transit-oriented regions. Although less than 1% of T/R trips involved a direct transfer to or from transit, one-third of all tours containing T/R also included transit. However, at the same time, 40% of T/R-containing tours also involved auto trip(s). Overall, this study reveals the complex relationships among their underlying sociodemographic characteristics, RS adoption and usage behavior, and daily tour patterns.

Selecting pharmacies for COVID-19 testing to ensure access.

Rapid diagnostic testing for COVID-19 is key to guiding social distancing orders and containing emerging disease clusters by contact tracing and isolation. However, communities throughout the US do not yet have adequate access to tests. Pharmacies are already engaged in testing, but there is capacity to greatly increase coverage. Using a facility location optimization model and willingness-to-travel estimates from US National Household Travel Survey data, we find that if COVID-19 testing became available in all US pharmacies, an estimated 94% of the US population would be willing to travel to obtain a test, if warranted. Whereas the largest chain provides high coverage in densely populated states, like Massachusetts, Rhode Island, New Jersey, and Connecticut, independent pharmacies would be required for sufficient coverage in Montana, South Dakota, and Wyoming. If only 1,000 ZIP code areas for pharmacies in the US are selected to provide testing, judicious selection, using our optimization model, provides estimated access to 29 million more people than selecting pharmacies simply based on population density.Supplementary InformationThe online version contains supplementary material available at (10.1007/s10729-020-09538-w).

Daily electric vehicle charging load profiles considering demographics of vehicle users

Travel pattern of an electric vehicle (EV) user and the accuracy of their probability distribution models are the key factors affecting the simulation and prediction of EV charging load. Most of the existing works utilized the travel data for all kinds of populations and ignored the influence of people social attributes on their travel pattern, which deteriorates the accuracy of the charging load model. This paper demonstrates that the daily EV charging load profiles vary with different demographic and social attributes by presenting a refined EV charging load simulation method considering people’s demographics and social characteristics, e.g. gender, age, education level. First, to improve the fitting accuracy of people travel pattern, new probabilistic models of many defined spatial–temporal variables are established under refined conditions (i.e. location, day type, etc.). Second, additional factors (i.e. charging preference, power consumption rate, etc.) are included to simulate the daily profile of EV charging load based on the refined probabilistic models and Monte Carlo algorithm. Data from the US National Household Travel Survey are used to validate the proposed method. The results show that the user's demographic and social attributes have a considerable effect on the magnitude and peak time of the EV charging load profile, particularly for workdays and workplace. The proposed probabilistic models can improve the accuracy of the data fitting and the charging load simulation.

Trends in Taxi Use and the Advent of Ridehailing, 1995–2017: Evidence from the US National Household Travel Survey

The advent of ridehailing services such as Uber and Lyft has expanded for-hire vehicle travel. We use data from the 2017 National Household Travel Survey (NHTS) to investigate the extent of this expansion in the United States. We report changes in the for-hire vehicle market since ridehailing services became available and statistically estimate the determinants of ridehailing use. From 2009–2017, the for-hire vehicle market share doubled. While for-hire vehicles still only account for 0.5% of all trips, the percent of all Americans who use ridehailing in any given month is nearly 10%. Within the for-hire vehicle market, this trend of growth has not been uniformly distributed across demographic groups or geographies; it has been greater in mid-sized and large cities, and among younger individuals and wealthier households. This suggests that understanding the equity implications of ridehailing is an important avenue for research. Multivariate analysis provides evidence that both transit and nonmotorized transport use are correlated with ridehailing use, that ridehailing has a negative relationship with vehicle ownership, and that residents of denser areas have higher ridehailing use. Given the rapid growth of ridehailing, it has become important for cities to include for-hire vehicles in their planning going forward. These NHTS data provide a starting point, but more detailed and frequent data collection is needed to fully understand this many-faceted, rapidly-changing market.

Optimizing copious activity type classes based on classification accuracy and entropy retention

Despite the advantages, big transport data are characterized by a considerable disadvantage as well. Personal and activity-travel information are often lacking, making it necessary to deduce this information with data mining techniques.However, some studies predict many unique activity type classes (ATCs), while others merge multiple activity types into larger ATCs. This action enhances the activity inference estimation, but destroys important activity information. Previous studies do not provide a strong justification for this practice. An objectively optimized set of ATCs, balancing model prediction accuracy and preserving activity information from the original data, becomes essential.Previous research developed a classification methodology in which the optimal set of ATCs was identified by analyzing all possible ATC combinations. However, this approach is practically impossible in a finite amount of time for e.g. the US National Household Travel Survey (NHTS) 2009 data set, which comprises 36 ATCs (home activity excluded), since there would be 3.82⋅1030 unique combinations (an exponential increase).The aim of this paper is to optimize which original ATCs should be grouped into a new class, and this for data sets for which it is impossible or impractical to simply calculate all ATC combinations. The proposed method defines an optimization parameter U (based on classification accuracy and information retention) which is maximized in an iterative local search algorithm. The optimal set of ATCs for the NHTS 2009 data set was determined. A comparison finds that this optimum is considerably better than many expert opinion activity type classification systems. Convergence was confirmed and large performance gains were found.

The millennial question: Changes in travel behaviour or changes in survey behaviour?

Abstract A great deal has been written in recent years in the travel behaviour research literature about “peak travel” or “peak car”—particularly with respect to the so-called “Millennials”, young adults currently between the ages of 16 and 35. The paper reports and extends the results of longitudinal multivariate regression analysis carried out on the US National Household Travel Survey data from 1995, 2001 and 2009. The results show robust evidence for a decline in auto miles traveled between 2001 and 2009 that cannot be explained by demographic, socio-economic or land use variables or by pure trend effects. The paper then provides a discussion of whether such changes in travel behaviour over time observed from surveys could be due to changes in the survey methods or the way in which different age groups respond to specific types of surveys. There is evidence that young adults have the highest non-response bias (underreporting of trips) for diary-based surveys, but it is not clear if this bias has been growing over time. The paper closes with some suggestions for further data collection.

The association between graduated driver licensing laws and travel behaviors among adolescents: an analysis of US National Household Travel Surveys.

BackgroundYoung novice drivers have crash rates higher than any other age group. To address this problem, graduated driver licensing (GDL) laws have been implemented in the United States to require an extended learner permit phase, and create night time driving or passenger restrictions for adolescent drivers. GDL allows adolescents to gain experience driving under low-risk conditions with the aim of reducing crashes. The restricted driving might increase riding with parents or on buses, which might be safer, or walking or biking, which might be more dangerous. We examined whether GDL increases non-driver travels, and whether it reduces total travels combining drivers and non-drivers.MethodsWe used data from the US National Household Travel Survey for the years 1995–1996, 2001–2002, and 2008–2009 to estimate the adjusted ratio for the number of trips and trip kilometers made by persons exposed to a GDL law, compared with those not exposed.ResultsAdolescents aged 16 years had fewer trips and kilometers as drivers when exposed to a GDL law: ratio 0.84 (95 % confidence interval (CI) 0.71, 1.00) for trips; 0.79 (0.63, 0.98) for kilometers. For adolescents aged 17 years, the trip ratio was 0.94 (0.83, 1.07) and the kilometers ratio 0.80 (0.63, 1.03). There was little association between GDL laws and trips or kilometers traveled by other methods: ratio 1.03 for trips and 1.00 for kilometers for age 16 years, 0.94 for trips and 1.07 for kilometers for age 17.ConclusionsIf these associations are causal, GDL laws reduced driving kilometers by about 20 % for 16 and 17 year olds, and reduced the number of driving trips by 16 % among 16 year olds. GDL laws showed little relationship with trips by other methods.

US vehicle fuel-efficiency choices: demographic, behavioral, and cultural factors

In the United States, a large proportion of greenhouse gas emissions are from personal vehicle use, and there are large differences in emissions between the most and least efficient vehicles. We use data from 2009 US National Household Travel Survey to characterize consumers who drive the most and least efficient vehicles. We calculate personal fleet efficiency based on all vehicles owned, reflecting the behavioral aspect of vehicle-use choices by owners. Though higher incomes provide the means to purchase more efficient vehicles, we find those with higher incomes more likely to drive low-efficiency vehicles. Higher education levels correlate with greater vehicle efficiency. While some efficiency differences are based on consumer characteristics such as family size, we find large differences that we attribute mostly to cultural preferences as observed by race, region, etc. From a policy perspective, these differences are of particular interest since they may be amenable to change.

Estimating Active Transportation Behaviors to Support Health Impact Assessment in the United States.

Health impact assessment (HIA) has been promoted as a means to encourage transportation and city planners to incorporate health considerations into their decision-making. Ideally, HIAs would include quantitative estimates of the population health effects of alternative planning scenarios, such as scenarios with and without infrastructure to support walking and cycling. However, the lack of baseline estimates of time spent walking or biking for transportation (together known as “active transportation”), which are critically related to health, often prevents planners from developing such quantitative estimates. To address this gap, we use data from the 2009 US National Household Travel Survey to develop a statistical model that estimates baseline time spent walking and biking as a function of the type of transportation used to commute to work along with demographic and built environment variables. We validate the model using survey data from the Raleigh–Durham–Chapel Hill, NC, USA, metropolitan area. We illustrate how the validated model could be used to support transportation-related HIAs by estimating the potential health benefits of built environment modifications that support walking and cycling. Our statistical model estimates that on average, individuals who commute on foot spend an additional 19.8 (95% CI 16.9–23.2) minutes per day walking compared to automobile commuters. Public transit riders walk an additional 5.0 (95% CI 3.5–6.4) minutes per day compared to automobile commuters. Bicycle commuters cycle for an additional 28.0 (95% CI 17.5–38.1) minutes per day compared to automobile commuters. The statistical model was able to predict observed transportation physical activity in the Raleigh–Durham–Chapel Hill region to within 0.5 MET-hours per day (equivalent to about 9 min of daily walking time) for 83% of observations. Across the Raleigh–Durham–Chapel Hill region, an estimated 38 (95% CI 15–59) premature deaths potentially could be avoided if the entire population walked 37.4 min per week for transportation (the amount of transportation walking observed in previous US studies of walkable neighborhoods). The approach developed here is useful both for estimating baseline behaviors in transportation HIAs and for comparing the magnitude of risks associated with physical inactivity to other competing health risks in urban areas.

Individual trip rate transferability analysis based on a decision tree approach

ABSTRACTTransferring trip rates to areas without local survey data is a common practice which is typically performed in an ad hoc fashion using household-based cross-classification tables. This paper applies a rule-based decision tree method to develop individual-level trip generation models for eight different trip purposes as defined in the US National Household Travel Survey in addition to daily vehicle miles traveled. For each trip purpose, the models are obtained by finding the best fitted statistical distribution to each of the final decision tree clusters while considering the correlation between the trip rates for other trip purposes. The rule-based models are sensitive to changes in demographics. The performance of the models is then tested and validated in a transferability application to the Phoenix Metropolitan Region. These models can be employed in a disaggregate microsimulation framework to generate trips with different purposes at the individual or household level.