Understanding Car Ownership Decisions: Evidence from an Emerging Central Business District in Mumbai, India

Assessment of urban transportation pricing policies with incorporation of unobserved heterogeneity

The secret of greatness is simple: do better work than any other man in your field – and keep on doing it. (Wilfred A. Peterson) Introduction The choice modeler has available a number of econometric models. Traditionally, the more common models applied to choice data are the multinomial logit (MNL) and nested logit (NL) models. Increasingly, however, choice modelers are estimating the mixed logit (ML) or random parameters logit model. In Chapter 4, we outlined the theory behind this class of models. In this chapter we estimate a range of ML models using Nlogit, including recent developments in scaled mixed logit (or generalized mixed logit). As with Chapters 11 and 13 (MNL model) and Chapter 14 (NL model), we explain in detail the commands necessary to estimate ML models as well as the interpretation of the output generated by Nlogit. An understanding of the theory behind the ML model is presented in Chapter 4; however we anticipate that in reading this chapter you will have a better understanding of the model, at least from an empirical standpoint. The mixed logit model basic commands The ML model syntax commands build on the commands of the MNL model discussed in Chapter 11. We begin with the basic ML syntax command, building upon this in later sections as we add to the complexity of the ML model.

In order to further improve the accuracy of evaluation models on value of a statistical life (VOSL) in road traffic, four mixed logit (ML) models of route-choice with truncated normal distribution and lognormal distribution were used to construct VOSL models. A route-choice questionnaire was designed by the stated choice method, and the traffic survey was carried out in Dalian with survey data obtained. Monte Carlo simulation algorithm was used to calibrate parameters by 150 simulations, and the 4 ML models were analyzed comparatively. Finally, the VOSL estimate of private drivers in Dalian and its distribution function were obtained. The research results indicate: ML model with truncated normal distribution has \(\overline{\rho}^{2}\) of 0.1516 and hit ratio of 70.42%, which has a lower accuracy. 3 ML models with lognormal distribution have a high accuracy, whose \(\overline{\rho}^{2}\) are all between (0.2–0.4) and hit ratios all above 80%. The 4th ML model whose parameters of fatal risk and travel cost obeying lognormal distribution simultaneously has the highest accuracy, with the greatest \(\overline{\rho}^{2}\) (0.2534) and the highest hit ratio (84.76%). VOSL in road traffic based on the 4th ML model obeys lognormal distribution with parameters (2.0622, 0.67402) with the mathematical expectation of 986,840 RMB. The maximum probability is 9.45% when VOSL is 500,000 RMB.

Effect of Distributional Assumption of Random Parameters of Mixed Logit Model on Willingness-to-Pay Values

A computationally attractive model for the analysis of conjoint choice experiments is the mixed multinomial logit model, a multinomial logit model in which it is assumed that the coefficients follow a (normal) distribution across subjects. This model offers the advantage over the standard multinomial logit model of accommodating heterogeneity in the coefficients of the choice model across subjects, a topic that has received considerable interest recently in the marketing literature. With the advent of such powerful models, the conjoint choice design deserves increased attention as well. Unfortunately, if one wants to apply the mixed logit model to the analysis of conjoint choice experiments, the problem arises that nothing is known about the efficiency of designs based on the standard logit for parameters of the mixed logit. The development of designs that are optimal for mixed logit models or other random effects models has not been previously addressed and is the topic of this paper. The development of efficient designs requires the evaluation of the information matrix of the mixed multinomial logit model. We derive an expression for the information matrix for that purpose. The information matrix of the mixed logit model does not have closed form, since it involves integration over the distribution of the random coefficients. In evaluating it we approximate the integrals through repeated samples from the multivariate normal distribution of the coefficients. Since the information matrix is not a scalar we use the determinant scaled by its dimension as a measure of design efficiency. This enables us to apply heuristic search algorithms to explore the design space for highly efficient designs. We build on previously published heuristics based on relabeling, swapping, and cycling of the attribute levels in the design. Designs with a base alternative are commonly used and considered to be important in conjoint choice analysis, since they provide a way to compare the utilities of pro- files in different choice sets. A base alternative is a product profile that is included in all choice sets of a design. There are several types of base alternatives, examples being a socalled outside alternative or an alternative constructed from the attribute levels in the design itself. We extend our design construction procedures for mixed logit models to include designs with a base alternative and investigate and compare four design classes: designs with two alternatives, with two alternatives plus a base alternative, and designs with three and with four alternatives. Our study provides compelling evidence that each of these mixed logit designs provide more efficient parameter estimates for the mixed logit model than their standard logit counterparts and yield higher predictive validity. As compared to designs with two alternatives, designs that include a base alternative are more robust to deviations from the parameter values assumed in the designs, while that robustness is even higher for designs with three and four alternatives, even if those have 33% and 50% less choice sets, respectively. Those designs yield higher efficiency and better predictive validity at lower burden to the respondent. It is noteworthy that our “best” choice designs, the 3- and 4-alternative designs, resulted not only in a substantial improvement in efficiency over the standard logit design but also in an expected predictive validity that is over 50% higher in most cases, a number that pales the increases in predictive validity achieved by refined model specifications.

Determining mode shift elasticity based on household income and travel cost

The operation of minibuses as public transport modes worldwide is considered as a both efficient and effective way to meet the mobility needs of special population groups in urban areas. Minibus services are more flexible in terms of routing and scheduling as well as more affordable compared to regular bus public transport. This paper investigates the potential introduction and the characteristics of a minibus transit service in the Municipality of Kalamaria (~90,000 pop.) in Thessaloniki Greater Area, Greece. The research is based on a joint revealed and stated preference survey, which took place in 2015 and was addressed to the citizens of the Municipality, where such a minibus service was formerly active. The revealed preference survey mainly explored the current travel behavior of citizens and their level of acceptance for such a service in case of its reopening. The stated preference survey investigated the intentions and travel preferences of the respondents under different scenarios of travel time, service frequency and travel cost. A total of 140 questionnaires was collected and analyzed using descriptive and inferential statistics. Causal analysis was conducted to reveal interrelations of respondents’ willingness to use the service depending on the above attributes. According to the research findings, the perceived “quality of service” is strongly associated with safety, reliability and comfort of the transport mode used. Additionally, travel time is considered more important than cost. Although respondents seem to prefer a free of charge service, it was found that Willingness-to-Pay for minibus services ranges around 0.5 euros.

Exploring the role of social capital influence variables on travel behaviour

Estimating travelers’ trade-off between travel cost and other attributes accurately is significant for transport planning and project investment appraisal. This study introduces the nested logit (NL) model that accounts for joint choice of route and transport modes. A willingness to pay (WTP) method combined with stated preference (SP) survey is presented to estimate the value of travel time savings (VTTS) under multifactor influence. The results confirm that various attributes (i.e. travel time and cost, walking and waiting time, frequency of left turns, income level, age and car ownership) do play a part in the estimation of the value of travel time savings. It is found that the VTTS for car is highest among the WTP values of interest. Furthermore, non-car owners are more sensitive to their income and the level of service of public transportation, and value their WTP for travel time and walking and waiting time lower compared to car owners.

Exploring micromobility mode preferences for last-mile trips from subway stations

Modeling departure time choice of metro passengers with a smart corrected mixed logit model - A case study in Beijing

Discrete response models have been applied to many fields such as resource management, transportation studies, commodity brand studies, and recreation studies. The mixed logit model (MXL) is interesting because it is flexible, in a sense that it is a generalization of all discrete response models based on random utility maximization. The MXL can capture the random heterogeneous preference of the consumers, which is its unique strength among discrete response models. Typically, the ultimate goal of recreation studies is to obtain economic benefits related to changing site attributes or availability. The literature states that the welfare effects obtained by using the MXL are significantly affected by the interpretation of its random terms. One interpretation denotes the randomness as capturing taste differences among the consumers; another denotes the random terms as capturing measurement errors. The literature finds that the welfare effects the welfare effects based on the first interpretation are all substantially larger than those the welfare effects based on the second interpretation. This study uses a different data set. With more sites and under more distributions of the random parameters, I find that the interpretation of its random terms sometimes significantly affects the welfare effects. However, the welfare effects based on the first interpretation is not always larger than the welfare effects based on the second interpretation. This study also uses empirical data to illustrate that, when the travel cost is a random parameter, the welfare effects based on the first interpretation is significantly biased. A third formula based on the same interpretation is appropriate. These findings help understanding the welfare measurement when using the MXL. Therefore, it helps the application of the MXL in the study of natural recreational resource management as well as in other fields.

Vehicles are a major invention having significant impact on human progress.However, increasing private vehicle usage can cause many threats to the society like congestion, depleting environmental quality etc. The study is taken up to identify the factors influencing usage of private and public modes, with the help of mode choice models. The influence of psychological factors based on the ‘Theory of Planned Behaviour’ were analysed in the study.The variables considered for the study include service attributes like travel cost, travel time, waiting time, comfort and latent variables like ‘attitude’, ‘subjective norm’, ‘perceived behavioural control’ etc. Revealed preference survey was adopted to know the sociodemographic details and a stated preference survey was adopted to study the mode shift of people to a better public transit facility.Travel time, waiting time and travel cost had negative influence in the utility of new public transit mode.Comfort and perceived behavioural control had positive influence in utility of a new public transit service.The study was useful in developing a better understanding of mode choice behaviour and developing insights for policy formulations for improving the public transit.

In this paper, logit models as well as mixed multinomial logit models(MMNL) were established to model traffic split. The properties of Johnson distribution bounded on both sides distribution was analyzed fisrt and it was found that Johnson distribution which are bounded on both sides can describe uniform preference factors better than normal and log- normal distribution. Based the residents' daily travel survey of Hefei city, a logit model and two distribution mixed logit model(MMNLl and MMNL2) with different restriction condition were established to model traffic mode split of a certain area, and the results indicats that MMNL2 is more fitable to explain the original data and more reasonable to describe uniform preference factors such as travel time. The sensitivity of the mixed logit model to transit travel time factor was also discussed in the paper, and the results shows that the implementation of bus priority policy make the passenger flow share of the bus get some increase which mainly comes from the bike users. Furthermore, mixed logit model is more reasonable to model traffic mode split due to its reasonable to describe uniform preference factors and no possession of IIA limitation.

In China, a developing country, the car ownership level is much lower than that in developed countries, but transportation policies have been implemented to discourage car ownership and mitigate traffic congestion. However, car ownership (considered as car availability in this paper, meaning that an individual has access to a household private car) may influence travelers’ well-being. To highlight the interrelation between car ownership and travelers’ well-being, this paper develops a probit-based discrete-continuous model to analyze the relationship between car ownership and the duration of commuters’ three major non-work outdoor activities (Act1: shopping and dining; Act2: leisure and entertainment; and Act3: visiting relatives or friends) in Xiaoshan District, Hangzhou, China. Empirical results indicate strong effects of individual and household socio-demographics, built environment attributes, and work-related characteristics on the car ownership decision and the duration of three non-work activities. The analysis shows positive correlations in unobserved factors between the car ownership decision and the duration of Acts1–3, indicating a mutually promotive relationship. Similarly, negative correlations among the duration of Acts1–3 show that non-work activities’ duration is mutually substitutive. These findings will help to better understand commuters’ car ownership decisions and non-work outdoor activity behavior restricted by fixed work schedules in developing countries, which can, in turn, better evaluate the impact of transportation policies (such as car ownership restriction) on travel demand as well as well-being, and provide decision support for the formulation of transportation policies.