For some decades now, transport researchers have put considerable efforts into developing what is called activity-based approaches for modelling urban travel demand. The basic idea is that travel demand is derived from people's desires to take part in different activities. In particular, the interrelationships among different activities with respect to temporal and spatial constraints are in focus. It means that such models treat the activities and the travelling of the households with respect to where and when the activities can be carried out and how they may be scheduled, given characteristics of the households and potential opportunities, the transport networks and various institutional constraints. We discuss what demands we see on future travel demand models, with a focus on urban analysis. This discussion is somewhat biased towards what role activity-based models could play in meeting these demands. We then review in some detail three prominent and distinctly different representatives of operational activity-based models to give an indication of what new modelling possibilities they offer. Theoretical appeal, empirical validity, usefulness for planning, need for data and easiness of implementation are discussed. In the final section we draw some conclusions about the prospects of these models and of their descendants.

Swedish noise values are today based on a study concerning hedonic price values. The valuation of the average noise levels assumes that there is no difference between different noise profiles, i.e. how the noise level varies throughout the day. In order to be able to take appropriate measures towards noise problems, it is important to know to what extent the noise profile matters. The project aims at researching how the noise value depends on the road traffic noise profile. It further aims at studying how values are affected by the type of activity in which the noise disturbance occurs (being out in the garden/try to sleep in the bedroom). Our method is to use stated preference techniques (pair wise choices) to elicit noise profile values. Respondents listen to and evaluate noise profile where we vary the level of background noise, frequency of noise events and the noise level of the events. This paper also discusses possibilities to estimate willingness to pay for changes in different components of noise profiles.

The subject of this thesis concerns pricing the use of transport infrastructure. We are empirically investigating the relationship between railway traffic volumes and infrastructure management costs. More specifically, we are interested in estimating the change in infrastructure management costs from marginal variations in traffic volumes, i.e. to estimate the marginal cost of railway infrastructure wear and tear. Both Europe and Sweden have moved towards a marginal cost based transport pricing policy, thus driving the need for more empirical work on rail infrastructure costs to underpin the level of a wear and tear charge. The thesis consists of five papers. In paper I, the data situation for planning railway maintenance and renewal is surveyed internationally. The survey indicates that most infrastructure managers are still in the data gathering phase, rather than ready to use modern computerised planning tools to make sound decisions in the field of maintenance and renewal. In paper II, we investigate the data situation for infrastructure cost analysis in Sweden. A panel data set that consists of cost, traffic and infrastructure information is created. The data covers 1999-2002 and contains almost 190 annual observations. Three main cost categories are identified; infrastructure operation, maintenance and renewal. This data is used for estimations of cost functions in paper II, III and V. Econometric techniques are applied for this purpose, with several different model specifications. In paper II, the method of pooled ordinary least squares (POLS) is applied. In paper III, we turn to unobserved effects models to exploit data heterogeneity. Finally in paper V, a dynamic generalised method of moments estimator is used to explore a potential dynamic cost dependency. The main findings are that the POLS approach, which has been used in similar studies in Europe recently, is rejected in favour of fixed effects estimation for this data. Furthermore, we also reject the idea of regression analysis to capture marginal rail renewal costs. In paper IV, we suggest an analytical expression combined with survival analysis of rail ages to estimate marginal renewal costs. We derive elasticities with respect to output as well as marginal costs for the different cost categories, and find that the current charge for wear and tear in Sweden is well below these new estimates. This opens up for increased, marginal cost based rail infrastructure wear and tear charges, which would reduce the financial burden on Swedish tax-payers.

New railway legislation in Sweden has increased the need for transparent access charges on the Swedish railway network. We estimate cost functions for infrastructure operation, maintenance and renewal in the Swedish national railway network, using unobserved effects models and calculate marginal costs for railway infrastructure wear and tear. We find evidence of unobserved fixed effects at a track section level for infrastructure operation and maintenance costs. The estimated weighted average marginal infrastructure operation cost is SEK 0.12 per train kilometre and the estimated marginal maintenance cost is SEK 0.0073 per gross tonne kilometre. Altogether, the results indicate that the current charge for railway infrastructure wear and tear in Sweden is below marginal cost.

The separation of infrastructure and train operations in Europe's railway sector has increased the importance of a transparent policy for pricing infrastructure use. This paper reviews the quality of data on the Swedish railway network to allow estimation of the marginal costs of railway track wear and tear. Information about costs, traffic, and infrastructure is examined and assessed in light of an econometric modeling approach. The review shows that the basic data necessary to estimate an econometric cost function are not always at hand. In particular, no extensive time series data are available, and there are instances of partially missing data. However, some problems can be overcome through the use of data imputation techniques. A database covering 749 track section observations from 1999 to 2002 was created. Cost functions were estimated via pooled ordinary least squares, and marginal costs were calculated. Separate charging regimes for operational and maintenance costs are suggested, but more research is needed before renewal costs can be recommended for inclusion.

In this paper, dynamic aspects of railway infrastructure operation and maintenance costs in Sweden are explored. Econometric cost functions are estimated to check the robustness of previous marginal cost estimates by introducing lags and leads of both dependent and independent variables. We find support for a forward-looking behaviour within the Swedish National Rail Administration (Banverket) as both infrastructure operation and maintenance costs are reduced prior to a major renewal. There are also indications of both lagged traffic and costs affecting the cost structure.

In this paper, renewal costs for railway tracks are investigated using survival analysis. The purpose is to derive the effect from increased traffic volumes on rail renewal cycle lengths and to calculate associated marginal costs. A flow sample of censored data containing almost 1 500 observations on the Swedish main railway network is used. We specify Weibull accelerated failure time regression models, and estimate deterioration elasticities for total tonnage as well as for passenger and freight tonnages separately. Marginal costs are calculated as a change in present values of renewal costs from premature renewal following increased traffic volumes.

Vulnerability, exposure and criticality in various infrastructures are issues that have been more explicitly looked into in recent years. However, road vulnerability as such has not been in focus for very long, despite the fundamental importance of our road networks in everyday life, as well as in crisis evacuation situations. Consequently, network reliability in transport modelling is an important and growing field of research (Lam 1999). The connection between reliability, vulnerability and other related concepts are discussed in Berdica (2002), with the main proposition that vulnerability analysis of road networks should be regarded as an overall framework, within which different transport studies can be performed to describe how well our transport systems function when exposed to different kinds of disturbances. Following that approach, this paper presents the results from a model-based case study, performed with the overall objective to study how vulnerable the Stockholm road network is in different respects. More specifically it is built up around three main questions:

1.      How do interruptions of different critical links affect the system and how important are these links in relation to one another?

2.      How is the network performance affected by general capacity reductions and possible prioritisation of a sub-network?

3.      How is the system affected by traffic demand variations, i.e. how close to its capacity limit does the system operate?

Congestion charges were introduced in Stockholm in 2006, first as a trial followed by a referendum, then permanently from 2007. This paper discusses what conclusions can be drawn from the first four years of operation. We show that the traffic reduction caused by the charges’ has increased over time, once external factors are controlled for. Alternative-fuel vehicles are exempt from the charges, and this has substantially increased the sales of such vehicles. We discuss public and political acceptability, synthesizing recent research and Swedish experience, and conclude that objective and subjective effects as well as general environmental and political attitude played a role for the high levels of public support, while institutional reform and resolving power issues were necessary to gain political support. Finally, we briefly discuss implications for the transport planning process in general.

We estimate the value of time gains, different cycling environments and additional benefits in cost-benefit analysis of cycling investments. Cyclists’ value of travel time savings turns out to be high, considerably higher than the value of time savings on alternative modes. Cyclists also value other improvements highly, such as separated bicycle lanes. As to additional benefits of cycling improvements in the form of health and reduced car traffic, our results do not support the notion that these will be a significant part in a cost-benefit analysis, contrary to some earlier studies and beliefs. As to health effects, cyclists seem to take these largely into account when making their travel choices, so it would be double-counting to add total health benefits to the analysis once the consumer surplus has been correctly calculated. As to reductions in car traffic, our results indicate that the cross-elasticity between car and cycle is low, and hence benefits from traffic reductions will be small. However, the valuations of improved cycling speeds and comfort are so high that it seems likely that improvements for cyclists are cost-effective compared to many other types of investments, without having to invoke second-order, indirect effects. In other words, the bicycle is perfectly able to be viewed as a competitive mode of travel, rather than primarily a means to achieve improved health or reduced car traffic.

This paper provides new evidence on the evolution of the value of travel time (VTT) over time and its relation to income based on two essentially identical Stated Choice experiments conducted at an interval of 13 years. The results indicate that the income elasticity of the VTT is not uniform over income but increasing in income. As a consequence, the average rate at which the VTT increases with income in the cross-sectional samples has itself increased between the two survey years and can be expected to increase further over time. The estimation results support the idea that the income elasticity of the value of time has remained constant at each real income level. This confirms that it is not so much the relationship between income and the value of travel time that has changed over time as it is the level and distribution of income in the samples that has changed.

In this paper we discuss Hicksian demand and compensating variation in the context of discrete choice. We first derive Hicksian choice probabilities and the distribution of the (random) expenditure function in the general case when the utilities are nonlinear in income. We subsequently derive exact and simple formulae for the expenditure and choice probabilities under price (policy) changes conditional on the initial utility level. This is of particular interest for welfare measurement because it enables the researcher to compute the distribution of compensating variation in a simple way. We also derive formulae for the joint distribution of expenditure, the choice before and after a policy change has been introduced.

It is widely recognised that congestion pricing could be an effective measure to solve environmental and congestion problems in urban areas—a reform that normally also would generate a net welfare surplus. Despite this the implementation of congestion pricing has been very slow. One reason for a low public and political acceptance could be that equity impacts have not been given enough concern. In studies of distributional impacts of congestion pricing it has often been claimed that the reform is regressive rather than progressive even if there are studies claiming the opposite. We develop a method for detailed, quantitative assessment of equity effects of road pricing and apply it to a real-world example, namely a proposed congestion-charging scheme for Stockholm. The method simultaneously takes into account differences in travel behaviour, in preferences (such as values of time) and in supply of travel possibilities (car ownership, public transport level-of-service etc.). We conclude that the two most important factors for the net impact of congestion pricing are the initial travel patterns and how revenues are used. Differences in these respects dwarf differences in other factors such as values of time. This is accentuated by the fact that the total collected charges are more than three times as large as the net benefits. With respect to different groups, we find that men, high-income groups and residents in the central parts of the city will be affected the most. If revenues are used for improving public transport, this will benefit women and low-income groups the most. If revenues are used for tax cuts, the net benefits will be about equal for men and women on the average, while it naturally will benefit high-income groups. Given that it is likely that the revenues will be used to some extent to improve the public transport system, we conclude that the proposed congestion-charging scheme for Stockholm is progressive rather than regressive.

Route choice modeling is complex. The number of alternative paths is often very large, while the paths are likely to share unobserved attributes which induces correlation. When modelling this, we face a trade-off between using models that are simple enough to handle many alternative paths while at the same time being able to handle correlation. There is a substantial ongoing research effort seeking to resolve this dilemma, so far with limited success. For these reasons the multinomial logit model (path size logit and c-logit proposed by Ben-Akiva and Bierliare, 1999, and Cascetta et. al., 1996, respectively) is widely used in spite of its known limitations.

The main purpose of this paper is to present and test a dynamic discrete choice approach for the estimation of the parameters of a route choice model. In the dynamic modeling approach, the individual is seen as taking sequential decisions on which link to choose, and the choices are made at the nodes in the network. The obvious advantage with this approach is that the choice set at every stage is quite small and well defined, while a correlation structure is naturally imposed among different paths, even if each sequential decision follows a multinomial logit model. From an econometric point of view, the link choice model can be a lot simper to deal with.

The utility maximising choice of path may be broken down into a sequence of link choices, where at each stage the individual considers the utility associated with downstream link choices accumulated into a value function. However, if we were to compute the value function associated with the available link choices at every stage, the complexity of the problem would be at least the same as the original path choice problem. An exact solution method to calculate the value function runs into the curse of dimensionality when solving a dynamic programming problem. Therefore, the computational burden may be prohibitive for large networks if one tries to solve the dynamic programming problem by brute force. This is probably why the sequential approach is not used for estimating route choice models in spite of having been around for many years (e.g., Dial, 1971).

However, it is not strictly necessary to solve the dynamic programming problem in order to estimate the parameters of the route choice model consistently. It is sufficient to find a suitable approximation to the value function. So the objective of this paper is to test whether it is possible to generate good predictors for the value function such that the parameters of the route choice model may be estimated on link choices rather than path choices. If this turns out to be possible, then both the econometric and computational complexity of route choice modelling may be dramatically reduced.

The paper therefore discusses the conditions under which the route choice model can be consistently estimated. We then test the approach using simulated data for a real network (Borlänge, Sweden), where route choice observations are generated using the exact model, i.e. solving the dynamic programming problem. This allows us to compare the exact value functions with the approximations. We show how the approximation can be defined using proxy variables such as direction and distance to destination. The paper concludes with a discussion on the use of the model for prediction (policy analysis) and related issues.

We derive the value of reliability in the scheduling of an activity of random duration, such as travel under congested conditions. Using a simple formulation of scheduling utility, we show that the maximal expected utility is linear in the mean and standard deviation of trip duration, regardless of the form of the standardised distribution of trip durations. This insight provides a unification of the scheduling model and models that include the standard deviation of trip duration directly as an argument in the cost or utility function. The results generalise approximately to the case where the mean and standard deviation of trip duration depend on the starting time. An empirical illustration is provided.

As evidenced by recent studies, travel time reliability forms a substantial portion of traveler preferences and of the costs or benefits of transport policies. This paper describes how new theoretical developments have provided the means to estimate the value of travel time reliability improvements under schedule-theoretic traveler preferences at the same time. In order to help bring these theories into practice, tools are needed to predict the relevant features of a travel time distribution. This paper has examined some empirical data for “mean lateness”, an important component of travel time reliability under scheduling behavior, and estimated a model of the absolute mean lateness that can be used in conjunction with a model of mean travel times to produce a complete valuation. The findings show that the model, while showing significant results, only partially explains variations in mean lateness, and can be subject to large bias in certain cases.

In recent years, there has been an increasing awareness that travel time reliability, apart from expected travel time, is an important component of cost-benefit analysis, especially during congested traffic conditions. A common measure of travel time reliability is standard deviation, and it has been shown that this is a theoretically sound measure under scheduling constraints, provided that the mean lateness factor is known. Hence, in applied cost-benefit analyses, one will need both the standard deviation and the mean lateness factor. These analyses would be particularly simple if the mean lateness were constant across time of day and for different routes chosen. A study was done to explore how the mean lateness varies and how its variations can be approximated. With the use of travel time measurements on individual links, it is shown how mean lateness varies considerably across time and space. It is shown that mean lateness exhibits a time-varying pattern depending on the characteristics of congestion on the link. It is also shown that the location of the link in the network is a significant determinant. The resulting model for mean lateness represents a considerable improvement over existing practice, where the mean lateness is implicitly assumed constant, yet a large portion of its variation remains unexplained. The model is useful for informing future research but is of less value for predicting the mean lateness in broad applied settings.

Predicting traffic of individual drivers' behavior is an important part of many transport relatedapplications, for example, traffic simulation and intelligent GPS navigation. Travelers have differenttrip purposes, habits, preferences, et cetera, and do not necessarily choose the shortest path. For thisreason, random utility models (RUM) are used and the topic of this talk is the estimation of such modelsbased on trip observations in real networks (revealed preferences data) and their application. Using theRUM framework, there are two main issues that need to be addressed; definition of alternatives andmodeling correlation among alternatives. The talk has the following outline. First a background to routechoice modeling (static uni-modal networks) is given and an application in the Swedish city of Borlängeis used as an example. Second, a sampling of alternatives approach for the definition of choice sets ispresented (joint work with Moshe Ben-Akiva and Michel Bierlaire). Finally, some future researchdirections are discussed.

This paper presents a new paradigm for choice set generation in the context of route choice model estimation. We assume that the choice sets contain all paths connecting each origin-destination pair. Although this is behaviorally questionable, we make this assumption in order to avoid bias in the econometric model. These sets are in general impossible to generate explicitly. Therefore, we propose an importance sampling approach to generate subsets of paths suitable for model estimation. Using only a subset of alternatives requires the path utilities to be corrected according to the sampling protocol in order to obtain unbiased parameter estimates. We derive such a sampling correction for the proposed algorithm. Estimating models based on samples of alternatives is straightforward for some types of models, in particular the multinomial logit (MNL) model. In order to apply MNL for route choice, the utilities should also be corrected to account for the correlation using, for instance, a path size (PS) formulation. We argue that the PS attribute should be computed based on the full choice set. Again, this is not feasible in general, and we propose a new version of the PS attribute derived from the sampling protocol, called Expanded PS. Numerical results based on synthetic data show that models including a sampling correction are remarkably better than the ones that do not. Moreover, the Expanded PS shows good results and outperforms models with the original PS formulation.

Gender differences in travel behaviors are relevant in addressing equity needs and a variety of policy issues.  In order to effectively design gender equity policy, whether transport-related or other, one must first understand the differences found in travel behaviors between genders and what the implications of those differences are.  This thesis presents two papers aimed at exploring a variety of travel dimensions on the gender level.

The first paper reports research on the travel pattern changes over 19 years in the Stockholm region and the influences of these changes.  The study is implemented by comparing two large-scale travel surveys carried out in Stockholm County from 1986-87 and 2004.  By use of statistical and econometric methods, this study analyzes changes in various dimensions of travel patterns, while controlling for changes in external factors, such as automobile ownership, residence location and employment status.  Particular focus is given to how travel in different lifecycle groups (defined by gender, age and household composition) has changed.  The findings show that while travel distance has increased overall, it increases at a higher rate for females than her male counterpart.  In other words, women's travel behavior has become more similar to men's which explains an overall trend increase in travel distance.

The second paper presents estimates of the value of commuting time obtained from a model of subjective life satisfaction in the context of a household.  The model uses a correlated error structure for the household dimension in the data and it is estimated as an ordered probit model.  Separate models are, furthermore, estimated for households with and without young children, to capture potential differences for females and males from having young children.  The results indicate that the presence of young children in a household has a significant effect on the influences of commuting time for partnered males and females.  In general when considering the influence of commuting time by gender, it is shown that there is little difference in disutility for females and males without young children but a higher disutility for commuting time for females than her male counterparts when there are young children in the household.  From these findings, it is concluded that the value of commuting time based on subjective life satisfaction is higher for females, particularly in households with young children.

This paper deals with route choice models capturing travelers' strategic behavior when adapting to revealed traffic conditions en route in a stochastic network. The strategic adaptive behavior is conceptualized as a routing policy, defined as a decision rule that maps from all possible revealed traffic conditions to the choices of next link out of decision nodes, given information access assumptions. In this paper, we use a specialized example where a variable message sign provides information about congestion status on outgoing links. We view the problem as choice under risk and present a routing policy choice model based on the cumulative prospect theory (CPT), where utility functions are nonlinear in probabilities and thus flexible attitudes toward risk can be captured. In order to illustrate the differences between routing policy and non-adaptive path choice models as well as differences between models based on expected utility (EU) theory and CPT, we estimate models based on synthetic data and compare them in terms of prediction results. There are large differences in path share predictions and the results demonstrate the flexibility of the CPT model to represent varying degrees of risk aversion and risk seeking depending on the outcome probabilities.

An important purpose of the road transport system is to allow people to commute in efficientand reliable ways. For various undesired reasons, however, link capacities are sometimes reduced or linksare closed completely. To assess and reduce the risk of such events, a key issue is to identify road linksthat are particularly important, i.e. roads where disruptions would have particularly severe consequences.This paper presents a method for incorporating user equity considerations into a road link importancemeasure. As a complement to measuring the total increase in vehicle travel time, we also measure thedisparity in the distribution among individual users. These two components are combined to form anequity-weighted importance measure. We study the properties of this measure both analytically and in afull-scale case study of the Swedish road network. A main result is that increasing the weight put on the equity aspect transfers importance from the main roads to smaller local roads. The use of the measure intransport policy and planning is discussed.

We investigate the significance of geography for road network vulnerability. By our definitions, a region isexposed if the consequences of a random road closure are severe for this region, and it is important if theconsequences of a road closure within the region are severe for the overall network traffic. In a case study ofthe Swedish road network we study the distribution of exposure and importance among municipalities andcounties. We find considerable regional inequity of exposure, in particular for long closure durations. Wealso identify network regions that are particularly important for the socio-economic efficiency of the roadtransport system. Furthermore, we have developed simple proxy variables that can be used as rough estimatesof exposure and importance and provide a better understanding of the factors underlying these measures. Thepolicy implications of the findings are discussed. 

Disruptions in the transport system can have severe impacts for affected individuals, businesses and the society as a whole. In this research, vulnerability is seen as the risk of unplanned system disruptions, with a focus on large, rare events. Vulnerability analysis aims to provide decision support regarding preventive and restorative actions, ideally as an integrated part of the planning process.The thesis specifically develops the methodology for vulnerability analysis of road networks and considers the effects of suddenly increased travel times and cancelled trips following road link closures. The major part consists of model-based studies of different aspects of vulnerability, in particular the dichotomy of system efficiency and user equity, applied to the Swedish road network. We introduce the concepts of link importance as the overall impact of closing a particular link, and regional exposure as the impact for individuals in a particular region of, e.g., a worst-case or an average-case scenario (Paper I). By construction, a link is important if the normal flow across it is high and/or the alternatives to this link are considerably worse, while a traveller is exposed if a link closure along her normal route is likely and/or the best alternative is considerably worse. Using regression analysis we show that these relationships can be generalized to municipalities and counties, so that geographical variations in vulnerability can be explained by variations in network density and travel patterns (Paper II). The relationship between overall impacts and user disparities are also analyzed for single link closures and is found to be negative, i.e., the most important links also have the most equal distribution of impacts among individuals (Paper III).In addition to links' roles for transport efficiency, the thesis considers their importance as rerouting alternatives when other links are disrupted (Paper IV). Such redundancy-important roads, found often to be running in parallel to highways with heavy traffic, may be warranted a higher standard than their typical use would suggest. We also study the vulnerability of the road network under area-covering disruptions, representing for example flooding, heavy snowfall or forest fires (Paper V). In contrast to single link failures, the impacts of this kind of events are largely determined by the population concentration, more precisely the travel demand within, in and out of the disrupted area itself, while the density of the road network is of small influence. Finally, the thesis approaches the issue of how to value the delays that are incurred by network disruptions and, using an activity-based modelling approach, we illustrate that these delay costs may be considerably higher than the ordinary value of time, in particular during the first few days after the event when travel conditions are uncertain (Paper VI).