ABSTRACT The aim of this review was to gain a better understanding of the factors driving road tr... more ABSTRACT The aim of this review was to gain a better understanding of the factors driving road transport demand for both passengers and freight in the UK by reviewing the literature on elasticity of road traffic demand, with a particular focus on key economic and demographic factors: namely, population growth, income growth and changes in fuel costs. The primary aim was to identify, by means of a rapid evidence assessment, what elasticity estimates were available in the literature with respect to these variables and, where evidence exists, how these elasticity values have changed over time, if indeed they have changed at all. The range of estimated fuel price elasticity values reported in the studies in this review is quite small (-0.1 to -0.5), although a variety of data types and methodologies were used. Fuel price elasticities will be expected to vary by distance, area type and trip purpose. For passenger transport, reported income elasticity values are predominately in the range 0.5 to 1.4. The evidence indicates that car ownership has a strong, positive, indirect effect on the income elasticity of demand. For freight transport, elasticity estimates of economic activity are mainly in the range 0.5 to 1.5 for an aggregate commodity sector but there the evidence suggests a much greater variation between sectors. The evidence on changes in fuel price and income elasticities of car demand over time is limited and for freight transport, the evidence is mixed. Much of the data for the UK on car traffic is rather old. This has implications for the use of elasticities in forecasting and strategic planning.
We discuss the development of predictive choice models that go beyond the random utility model in... more We discuss the development of predictive choice models that go beyond the random utility model in its narrowest formulation. Such approaches incorporate several elements of cognitive process that have been identified as important to the choice process, including strong dependence on history and context, perception formation, and latent constraints. A flexible and practical hybrid choice model is presented that integrates many types of discrete choice modeling methods, draws on different types of data, and allows for flexible disturbances and explicit modeling of latent psychological explanatory variables, heterogeneity, and latent segmentation. Both progress and challenges related to the development of the hybrid choice model are presented.
While the paradigm of utility maximisation has formed the basis of the majority of applications i... more While the paradigm of utility maximisation has formed the basis of the majority of applications in discrete choice modelling for over 40 years, its core assumptions have been questioned by work in both behavioural economics and mathematical psychology as well as more recently by developments in the RUM-oriented choice modelling community. This paper reviews the basic properties with a view to explaining the historical pre-eminence of utility maximisation and addresses the question of what departures from the paradigm may be necessary or wise in order to accommodate richer behavioural patterns. We find that many, though not all, of the behavioural traits discussed in the literature can be approximated sufficiently closely by a random utility framework, allowing analysts to retain the many advantages that such an approach possesses.
Transportation Research Part C: Emerging Technologies, 2017
Communication patterns are an integral component of activity patterns and the travel induced by t... more Communication patterns are an integral component of activity patterns and the travel induced by these activities. The present study aims to understand the determinants of the communication patterns (by the modes face-to-face, phone, e-mail and SMS) between people and their social network members. The aim is for this to eventually provide further insights into travel behaviour for social and leisure purposes. A social network perspective brings value to the study and modelling of activity patterns since leisure activities are influenced not only by traditional trip measures such as time and cost but also motivated extensively by the people involved in the activity. By using a multiple discrete-continuous extreme value model (Bhat 2005), we can investigate the means of communication chosen to interact with a given social network member (multiple discrete choices) and the frequency of interaction by each mode (treated as continuous) at the same time. The model also allows us to investigate satiation effects for different modes of communication. Our findings show that in spite of people having increasingly geographically widespread networks and more diverse communication technologies, a strong underlying preference for face-to-face contact remains. In contrast with some of the existing work, we show that travel-related variables at the ego level are less important than specific social determinants which can be considered while making use of social network data.
ABSTRACT In the last few years a large share of the choice modelling work in transport and other ... more ABSTRACT In the last few years a large share of the choice modelling work in transport and other fields has been based on the use of data in which surveyed individuals give multiple responses. While the availability and use of multiple choices per individual has a number of significant advantages, we cannot avoid the problem that responses from a given individual cannot be treated as independent. Advanced modelling frameworks, especially those accommodating random variations in preferences, allow for an explicit treatment of the repeated choice nature of the data, but important questions need to be addressed in terms of choice of random parameters, the joint distributions used for these parameters, and the sequence of calculation of their integrals in the log-likelihood function. We discuss three different approaches that avoid the use of advanced modelling approaches while ensuring that the estimated errors are not unduly affected by the repeated choice nature of the data. We illustrate some of the properties of these approaches with real and simulated data. In this context, we show how a well used error correction method, the sandwich estimator, can be easily extended to deal with panel effects in a seemingly very reliable manner. We also explain how to do this with well known software. As an aside, we illustrate the well known theorem that if any correlation across choice tasks if independent of the attributes of the alternatives, then a naive model such as MNL will produce consistent estimates. If on the other hand, the correlation is due for example to unobserved taste heterogeneity, then the MNL estimates will no longer be consistent.
The scaling approach is a statistical estimation method which allows for differences in the amoun... more The scaling approach is a statistical estimation method which allows for differences in the amount of unexplained variation in different types of data which can then be Used together in analysis. In recent years, this approach has been tested and recommended in the context of combining Stated Preference and Revealed Preference data. The paper provides a description of the approach and a historical overview. The scaling approach can also be used to identify systematic differences in the variance of choices within a single Stated Preference data set due to the way in which the hypothetical choice situations are presented or the responses are obtained. The paper presents the results of two case studies-one looking at rank order effect and the other at fatigue effect. Scale effects appear to exist in both cases: the amount of unexplained variance is shown to increase as rankings become lower, and as the number of pairwise choices completed becomes greater. The implications of these findings for the use of SP ranking tasks and repeated pairwise choice tasks are discussed.
The traditional 4-step transport model has used the origin-destination trip as the basic unit of ... more The traditional 4-step transport model has used the origin-destination trip as the basic unit of analysis. The leading state of practice for transport models has shifted to production-attraction or tour based models with the most advanced models nowadays being activity based models. Both tour and activity based models use a tour rather than a trip as the basic unit of analysis. Whilst there are many differences in these modelling techniques, it is relatively simple to analyse household travel survey data from both a trip and tour perspective. This paper takes travel data from the Sydney Household Travel Survey and analyses it using both a trip and a tour methodology. Comparisons are then made of the results from this analysis. Major differences include insights into non-home based travel, the proportion of journeys which involve deviations (intermediate stops) and the resulting additional kilometrage. Also, as most people start and finish the day at home, it is possible to examine t...
The multinomial probit model has long been used in transport applications; as the basis for mode-... more The multinomial probit model has long been used in transport applications; as the basis for mode-and routechoice in computing network flows, and in other choice contexts when estimating preference parameters. It is well known that computation of the probit choice probabilities presents a significant computational burden, since they are based on multivariate normal integrals. Various methods exist for computing these choice probabilities, though standard Monte Carlo is most commonly used. In this paper we compare two analytical approximation methods (Mendell-Elston and Solow-Joe) with three Monte Carlo approaches for computing probit choice probabilities. We systematically investigate a wide range of parameter settings and report on the accuracy and computational efficiency of each method. The findings suggest that the accuracy and efficiency of the optimal Mendell-Elston analytic approximation method offers great potential for wider use.
This paper introduces the concept of Primary Family Priority Time (PFPT), which represents a high... more This paper introduces the concept of Primary Family Priority Time (PFPT), which represents a high priority household decision to spend time together for in-home activities. PFPT is incorporated into a fully specified and operational activity based (AB) discrete choice model system for Copenhagen, called COMPAS, using the DaySim software platform. Structural tests and estimation results identify two important findings. First, PFPT belongs high in the model hierarchy, and second, strong interactions exist between PFPT and the other day level activity components of the model system. Forecasts are generated for a road pricing and congestion scenario by COMPAS and a comparison version of the model system that excludes PFPT. COMPAS with PFPT exhibits less mode changing and time-of-day shifting in response to pricing and congestion than the comparison version.
ABSTRACT The aim of this review was to gain a better understanding of the factors driving road tr... more ABSTRACT The aim of this review was to gain a better understanding of the factors driving road transport demand for both passengers and freight in the UK by reviewing the literature on elasticity of road traffic demand, with a particular focus on key economic and demographic factors: namely, population growth, income growth and changes in fuel costs. The primary aim was to identify, by means of a rapid evidence assessment, what elasticity estimates were available in the literature with respect to these variables and, where evidence exists, how these elasticity values have changed over time, if indeed they have changed at all. The range of estimated fuel price elasticity values reported in the studies in this review is quite small (-0.1 to -0.5), although a variety of data types and methodologies were used. Fuel price elasticities will be expected to vary by distance, area type and trip purpose. For passenger transport, reported income elasticity values are predominately in the range 0.5 to 1.4. The evidence indicates that car ownership has a strong, positive, indirect effect on the income elasticity of demand. For freight transport, elasticity estimates of economic activity are mainly in the range 0.5 to 1.5 for an aggregate commodity sector but there the evidence suggests a much greater variation between sectors. The evidence on changes in fuel price and income elasticities of car demand over time is limited and for freight transport, the evidence is mixed. Much of the data for the UK on car traffic is rather old. This has implications for the use of elasticities in forecasting and strategic planning.
We discuss the development of predictive choice models that go beyond the random utility model in... more We discuss the development of predictive choice models that go beyond the random utility model in its narrowest formulation. Such approaches incorporate several elements of cognitive process that have been identified as important to the choice process, including strong dependence on history and context, perception formation, and latent constraints. A flexible and practical hybrid choice model is presented that integrates many types of discrete choice modeling methods, draws on different types of data, and allows for flexible disturbances and explicit modeling of latent psychological explanatory variables, heterogeneity, and latent segmentation. Both progress and challenges related to the development of the hybrid choice model are presented.
While the paradigm of utility maximisation has formed the basis of the majority of applications i... more While the paradigm of utility maximisation has formed the basis of the majority of applications in discrete choice modelling for over 40 years, its core assumptions have been questioned by work in both behavioural economics and mathematical psychology as well as more recently by developments in the RUM-oriented choice modelling community. This paper reviews the basic properties with a view to explaining the historical pre-eminence of utility maximisation and addresses the question of what departures from the paradigm may be necessary or wise in order to accommodate richer behavioural patterns. We find that many, though not all, of the behavioural traits discussed in the literature can be approximated sufficiently closely by a random utility framework, allowing analysts to retain the many advantages that such an approach possesses.
Transportation Research Part C: Emerging Technologies, 2017
Communication patterns are an integral component of activity patterns and the travel induced by t... more Communication patterns are an integral component of activity patterns and the travel induced by these activities. The present study aims to understand the determinants of the communication patterns (by the modes face-to-face, phone, e-mail and SMS) between people and their social network members. The aim is for this to eventually provide further insights into travel behaviour for social and leisure purposes. A social network perspective brings value to the study and modelling of activity patterns since leisure activities are influenced not only by traditional trip measures such as time and cost but also motivated extensively by the people involved in the activity. By using a multiple discrete-continuous extreme value model (Bhat 2005), we can investigate the means of communication chosen to interact with a given social network member (multiple discrete choices) and the frequency of interaction by each mode (treated as continuous) at the same time. The model also allows us to investigate satiation effects for different modes of communication. Our findings show that in spite of people having increasingly geographically widespread networks and more diverse communication technologies, a strong underlying preference for face-to-face contact remains. In contrast with some of the existing work, we show that travel-related variables at the ego level are less important than specific social determinants which can be considered while making use of social network data.
ABSTRACT In the last few years a large share of the choice modelling work in transport and other ... more ABSTRACT In the last few years a large share of the choice modelling work in transport and other fields has been based on the use of data in which surveyed individuals give multiple responses. While the availability and use of multiple choices per individual has a number of significant advantages, we cannot avoid the problem that responses from a given individual cannot be treated as independent. Advanced modelling frameworks, especially those accommodating random variations in preferences, allow for an explicit treatment of the repeated choice nature of the data, but important questions need to be addressed in terms of choice of random parameters, the joint distributions used for these parameters, and the sequence of calculation of their integrals in the log-likelihood function. We discuss three different approaches that avoid the use of advanced modelling approaches while ensuring that the estimated errors are not unduly affected by the repeated choice nature of the data. We illustrate some of the properties of these approaches with real and simulated data. In this context, we show how a well used error correction method, the sandwich estimator, can be easily extended to deal with panel effects in a seemingly very reliable manner. We also explain how to do this with well known software. As an aside, we illustrate the well known theorem that if any correlation across choice tasks if independent of the attributes of the alternatives, then a naive model such as MNL will produce consistent estimates. If on the other hand, the correlation is due for example to unobserved taste heterogeneity, then the MNL estimates will no longer be consistent.
The scaling approach is a statistical estimation method which allows for differences in the amoun... more The scaling approach is a statistical estimation method which allows for differences in the amount of unexplained variation in different types of data which can then be Used together in analysis. In recent years, this approach has been tested and recommended in the context of combining Stated Preference and Revealed Preference data. The paper provides a description of the approach and a historical overview. The scaling approach can also be used to identify systematic differences in the variance of choices within a single Stated Preference data set due to the way in which the hypothetical choice situations are presented or the responses are obtained. The paper presents the results of two case studies-one looking at rank order effect and the other at fatigue effect. Scale effects appear to exist in both cases: the amount of unexplained variance is shown to increase as rankings become lower, and as the number of pairwise choices completed becomes greater. The implications of these findings for the use of SP ranking tasks and repeated pairwise choice tasks are discussed.
The traditional 4-step transport model has used the origin-destination trip as the basic unit of ... more The traditional 4-step transport model has used the origin-destination trip as the basic unit of analysis. The leading state of practice for transport models has shifted to production-attraction or tour based models with the most advanced models nowadays being activity based models. Both tour and activity based models use a tour rather than a trip as the basic unit of analysis. Whilst there are many differences in these modelling techniques, it is relatively simple to analyse household travel survey data from both a trip and tour perspective. This paper takes travel data from the Sydney Household Travel Survey and analyses it using both a trip and a tour methodology. Comparisons are then made of the results from this analysis. Major differences include insights into non-home based travel, the proportion of journeys which involve deviations (intermediate stops) and the resulting additional kilometrage. Also, as most people start and finish the day at home, it is possible to examine t...
The multinomial probit model has long been used in transport applications; as the basis for mode-... more The multinomial probit model has long been used in transport applications; as the basis for mode-and routechoice in computing network flows, and in other choice contexts when estimating preference parameters. It is well known that computation of the probit choice probabilities presents a significant computational burden, since they are based on multivariate normal integrals. Various methods exist for computing these choice probabilities, though standard Monte Carlo is most commonly used. In this paper we compare two analytical approximation methods (Mendell-Elston and Solow-Joe) with three Monte Carlo approaches for computing probit choice probabilities. We systematically investigate a wide range of parameter settings and report on the accuracy and computational efficiency of each method. The findings suggest that the accuracy and efficiency of the optimal Mendell-Elston analytic approximation method offers great potential for wider use.
This paper introduces the concept of Primary Family Priority Time (PFPT), which represents a high... more This paper introduces the concept of Primary Family Priority Time (PFPT), which represents a high priority household decision to spend time together for in-home activities. PFPT is incorporated into a fully specified and operational activity based (AB) discrete choice model system for Copenhagen, called COMPAS, using the DaySim software platform. Structural tests and estimation results identify two important findings. First, PFPT belongs high in the model hierarchy, and second, strong interactions exist between PFPT and the other day level activity components of the model system. Forecasts are generated for a road pricing and congestion scenario by COMPAS and a comparison version of the model system that excludes PFPT. COMPAS with PFPT exhibits less mode changing and time-of-day shifting in response to pricing and congestion than the comparison version.
Uploads
Papers by Andrew Daly