Search

A key issue in the development of China's growing megacities in the transport-related environmental costs due to rapid urban expansion. In light of this issue, the authors examine the impact of urban form on commuting patterns on the city fringe of Beijing. Based on household-survey data, the results of the analysis suggest that the forms of land use adopted in the suburbs have a significant impact on commuting distance when a worker's socioeconomic characteristics and the level of transport accessibility are taken into account. Sprawling expansion, characterized by a low degree of self-contained development and low-density land use, tends to increase the need for long-distance commuting to the central urban area. In contrast, compact urban development in the suburbs, particularly in the peripheral constellations of Beijing, would reduce the probability of long-distance commuting. The current trend in improving transport accessibility on the city fringe is likely to lead to further long-distance commuting. In particular, huge road projects could cause more traffic congestion in the centre. The findings suggest that land-development management on the city fringe could have significant implications with respect to containing the dramatic costs to the environment entailed by transportation in the context of the rapid process of motorization. Reducing travel needs through the integration of land use and transport-infrastructure provision is likely to be the key to sustainable urban expansion.

Transit-oriented development is shown to produce an appreciable ridership bonus in California. This is partly due to residential self-selection—that is, a lifestyle preference for transit-oriented living—as well as factors like employer-based policies that reduce free parking and automobile subsidies. Half-mile catchments of station areas appear to be indifference zones in the sense that residents generally ride transit regardless of local urban design attributes. Out-of-neighborhood attributes, like job accessibility and street connectivity at the destination, on the other hand, have a significant bearing on transit usage among station-area residents. The presence of self-selection, shown using nested logit modeling, underscores the importance of removing barriers to residential mobility so that households are able to sort themselves, via the marketplace, to locations well served by transit. Market-responsive zoning, flexible residential parking policies, location efficient mortgages, and adaptive reuse of parking lots are also promising tools for expanding the supply of transit-based housing.

Accessibility is often seen to be an important determinant of the location of economic activities. This paper focuses on the specification of accessibility indicators for modelling the location choices of offices, with particular application to the upcoming implementation of a high-speed railway line in the Netherlands. Potential accessibility indicators are formulated, whereby attention is given to the shape of the impedance function and to the role of competitive transport modes in a transport mode's accessibility effect. These indicators are then tested in a discrete choice model on the location of office employment. Finally the accessibility indicators are used to explore the effects of the upcoming domestic high-speed train services in the Netherlands. The analyses show that a Box–Cox impedance function performs best for this application and significantly better than the exponential and power functions. The derived potential accessibility indicators have much explanatory capability for location attractiveness at a regional level, but at an intraregional level connectivity measures become more influential. Finally, it has been found that the accessibility effect of the future high-speed train connection is larger for business travel than for commuting, the value of time of travellers being a dominant factor.

This paper reports the results of a stated preference study investigating the willingness-to-pay (WTP) of employees at the Amsterdam Zuidas for the presence of nonshopping and shopping facilities. The Amsterdam Zuidas area, surrounding the current train–metro–tram station Amsterdam Zuid World-Trade-Centre, is the largest multifunctional land-use project currently under development in the Netherlands. For nonshopping facilities, the results show that employees have the highest WTP for the presence of day-care centres and public transport facilities, and the lowest for public and recreation facilities. The average WTP for the presence of nonshopping facilities amounts to approximately €29 per month per employee. The WTP for the presence of shopping facilities is estimated at €25 per month per employee on average, and is in absolute value highest for supermarkets and lowest for flowershops and dry cleaners.

Empirical techniques adapted from ecosystem-resilience theory allow estimation of how public health and public order within the New York Metropolitan Region respond to perturbations driven by changes in policy or economic structure. This approach constitutes a rigorous methodology for health-impact assessment, providing a quantitative measure of the stability of the region. Contrary to entrenched cultural assumption, affluent suburban counties and impoverished central-city neighborhoods remain strongly linked through a probability-of-contact matrix well indexed by the daily journey to work. The public-health ecology of the New York Metropolitan Region is remarkably unstable, greatly amplifying perturbations through mechanisms analogous to positive feedback in mechanical systems, with the single greatest influence being the percentage of the population living in poverty. Given the New York region's overwhelming dominance of national patterns for the hierarchical diffusion of disease and disorder this result has significant policy implication. More explicitly, lowering the rate of poverty in and near New York City would markedly reduce the vulnerability of the United States to emerging infection.

In this paper we revisit the excess-commuting technique and its links with urban form. The uncertainty in measurement is highlighted, as are the problems relating to changes in excess commuting over time. The measure of the theoretical maximum commute is proposed and added to the traditional excess-commuting measure so that the use of both the minimum and maximum levels can capture the concept of commuting potential. This measure is what we call the ‘extended excess-commuting measure’. These concepts are tested through the use of a simulation exercise. As well as arguing for the inclusion of socioeconomic variables in analysis, we demonstrate that decentralisation in urban spatial structure can lead to either an increase or a decrease in average commuting distance. Some of the inconsistencies in the use of excess commuting can be reduced through the use of actual commutes together with the commuting range, as these factors in combination lead to a clearer understanding of commuting efficiency.

Much of the analysis to date on the topic of excess commuting and jobs–housing balance deals with total commuting flow, undifferentiated with respect to worker and job characteristics. In this paper we explicitly address the disaggregation issue in terms of job and worker heterogeneity and show how to incorporate such details into the analysis of excess commuting. The objectives of this paper are (1) to develop a trip-distribution model disaggregating journey-to-work data according to type of occupation in order to estimate actual commutes; (2) to develop a disaggregated version of a linear program to measure theoretical minimum and maximum commutes; and, (3) to verify variations in excess commuting and jobs–housing balance according to type of occupation. Results of actual trip-length distributions for each occupation vary from 3.72 to 5 miles for Boise, Idaho, and from 4.27 to 7.78 miles for Wichita, Kansas. Minimum commutes vary from 0.95 to 3.58 miles and from 1.5 to 3.79 miles for Boise and Wichita, respectively. These results imply nonuniform levels of excess commuting and jobs/workers ratios. The proposed models are expected to have a wide range of uses in measurement and assessment of empirical patterns of commuting. The scope of the disaggregation can be extended to other targets, such as different types of industry, household structure, income level, ethnic background, education level, transportation mode, and gender. Further dimensions of disaggregation can address spatial interactions of different socioeconomic groups in urban areas, and, more generally, contribute to exploring urban sprawl according to job characteristics and industries.

In this paper, two forms of local regression are employed in the analysis of relations between out-commuting distance and other socioeconomic variables in Northern Ireland. The two regression approaches used are moving window regression (MWR) and geographically weighted regression (GWR). For the first approach different window sizes are applied and changes in results assessed. For the second approach, a Gaussian kernel is used and its bandwidth varied. Seven independent variables are utilised, although a single variable (deprivation) provides the main analytical focus. Differences in results obtained with use of the two approaches are discussed. The relationship between window size or bandwidth size and observed spatial patterning is discussed and elucidated. The results support previous work that indicated severe limitations in using global regressions to examine relationships between socioeconomic variables. Also, the utility of comparing results obtained from MWR and GWR is assessed and the benefits of both approaches are outlined.

Population censuses are among the primary sources of information on directional flows of migrants between places of usual residence, and of commuters between places of dwelling and work in Great Britain. The use of census origin–destination counts has in the past been restricted by data complexities and the lack of user accessibility. In order to improve accessibility a new, web-based interface has been developed which is the core of a Census Interaction Data Service, launched in 2002. The authors explain the Web-based Interface to Census Interaction Data from a user perspective, focusing on various features of the system that enable queries to be constructed and subsets of flow data to be extracted and downloaded in a quick and easy manner. Examples are provided to demonstrate alternative types of query and to illustrate the value of these data sets in providing insights into interaction behaviour.

The measurement of accessibility and travel patterns in urban areas is described. The methodology presented includes graphical measures of physical accessibility, a numerical index of accessibility that is consistent with graphical measures, residents' accessibility weighted by transport availability, and the relationship between accessibility and travel behaviour. Some empirical results are presented for access and travel to male and female jobs in Sydney, with the use of data collected for the 1971 Census of Population and for the Sydney Area Transportation Study. The consequences of some alternative arrangements of land use and plans to improve public transport on residents' accessibility are investigated. It is argued that accessibility measures are a useful aid to planners and policymakers in the social evaluation of urban structure.

Energy requirements for the journey to work, resulting from alternative population and employment growth patterns in a rural area of England, are estimated and compared. Both the absolute and relative energy-efficiency of alternative land-use patterns are shown to depend on assumptions about future mobility, implying that ideal spatial structures cannot be defined without reference to other important factors affecting travel behaviour and life-styles. Current trends in the area are leading to the land-use patterns which consistently emerged as being energy intensive in the analysis.

In part 1 of this paper (Putman and Ducca, 1978) we described a method for calibrating urban residential models in the absence of trip-interchange data (that is, the ‘trip matrix’). In part 2 we describe the empirical results of applying this method to a number of different urban areas. The results obtained are quite consistent both with theoretical expectations and from one area to another.

This paper explores the relationship between spatial structure and spatial interaction at the intraurban level. To examine this relationship an experimental framework is designed based on the application of a doubly constrained entropy-type gravity model to journey-to-work data for twenty-four Canadian urban areas. The study demonstrates that distance-decay exponents are strongly influenced by geographic structure and the geometry of origins and destinations. As such, both the influence of map pattern and the friction of distance should be explicitly incorporated into spatial interaction models. The paper also explores the impact of city size and the nature of the economic base of the urban area upon distance-decay exponents.

This paper presents an exploratory model of the population redistribution process in metropolitan areas. A set of ten differential equations developed from logistic growth theory describes the dispersal of population from an urban area into two encompassing concentric zones. These outward flows give rise to changes in the commuting flows by mode as well as changes in the modal mix of journey-to-work trips within the urban area. The model is calibrated for Greater London and the South-East Standard Region of England between 1961 and 1976, and various possible pricing scenarios are tested to examine the population trajectories up to 1991. The results illustrate the feasibility of dynamic process-descriptive models in urban analysis, but also suggest an integrated programme of research and model-building as part of this development.

When an office activity moves voluntarily in response to incentives provided by public policy, there is a presumption that the move benefits the relocating activity. But other activities may have to travel farther for face-to-face contact with the relocating activity; the resulting increases in transportation costs should be included in the system-wide effects of the relocation policy. A numerical example demonstrates that increases in system-wide costs can outweigh the benefits to the relocating activity. Empirical calibration of the model requires some modification of the contact diaries developed by Thorngren, Törnqvist, and others. Diary information is needed from activities which initiate contact with the relocating activity. Diary information on transportation costs for face-to-face contacts and for the journey to work can be used to conduct simulations of alternative relocation policies.

A unified approach to the modelling of household motor vehicle ownership level, consumption of vehicle holdings, and vehicle use is described. The model system is founded on a utility theory of household behavior, and is implemented through probabilistic discrete choice models estimated from cross-sectional samples of households. Following the introduction of the model, an application involving forecasts of US motor vehicle sales through the mid-1980s is presented.

A theory of cognitive structure is presented which is potentially capable of providing a conceptual framework for the study of transport behaviour, and by making use of this theory and the models derived from it, problematic areas in attitude and attribute research in transportation are discussed. An exploratory empirical study is described which successfully tests some of the relationships inherent in the theory in a transport context. Significant correlations were observed between overall attitude towards use of a travel mode and the products of evaluation and belief strength summed over a short series of perceived outcomes of use of that mode. The content, strength, and associated evaluations of salient beliefs about use of a travel mode were monitored during a period of change in the conditions of the service. Predictable changes in belief structure were observed, demonstrating the sensitivity of the techniques used.

The twin objectives of this study are first, to establish an unbiased statistical framework for interpreting the errors in prediction associated with Wilson's entropy-maximising transportation model and second, to apply the framework to the interpretation of journey-to-work patterns on Merseyside in 1966. The statistical framework centres on various decompositions of the index of information gain, while the empirical work highlights some anomalous trip-making behaviour that seems to be a result of public planning decisions prior to 1966.

The ‘causal’ role of geographical location can be overplayed in spatial analysis. In Northern Ireland the distribution of the dominant religious groups may be at least as important. Two models to explain social—economic patterns in this region are suggested, one based on religious groups and one upon location, and attempts are made to examine their separate influence in the geography of journey-to-work flows. The results suggest that location is the more important factor behind patterns at the inter-local-authority scale of analysis chosen.

This paper contains an examination of the fundamental assumption underlying the use of accessibility indicators: that an individual's travel behavior is related to his or her location vis-à-vis the distribution of potential activity sites. First, the conceptual and measurement issues surrounding accessibility and its relationship to travel are reviewed; then, an access measure for individuals is formulated. Using data from the Uppsala (Sweden) Household Travel Survey and controlling for sex, automobile availability, and employment status, the authors explore the relationship between both home- and work-based accessibility and five aspects of an individual's travel: mode use, trip frequencies and travel distances for discretionary purposes, trip complexity, travel in conjunction with the journey to work, and size of the activity space. From the results it can be seen that although all of these travel characteristics are related to accessibility to some degree, the travel–accessibility relationship is not as strong as deductive formulations have implied. High accessibility levels are associated with higher proportions of travel by nonmotorized means, lower levels of automobile use, reduced travel distances for certain discretionary trip purposes, and smaller individual activity spaces. Furthermore, the density of activity sites around the workplace affects the distances travelled by employed people for discretionary purposes. Overall, accessibility level has a greater impact on mode use and travel distance than it does on discretionary trip frequency. This result was unexpected in light of the strong trip frequency–accessibility relationship posited frequently in the literature.