Activity-based Measures of Accessibility for Transportation Policy Analysis
WANG D, Eindhoven University of Technology and International Institute for Aerospace Survey and Earth Sciences and TIMMERMANS H, Eindhoven University of Technology, The Netherlands
As a key and basic concept in location and transportation planning, accessibility conceptualises the opportunity provided by the transport and land-use system for different types of people to engage in activities (Jones, 1981). Various measures of accessi
As a key and basic concept in location and transportation planning, accessibility conceptualises the opportunity provided by the transport and land-use system for different types of people to engage in activities (Jones, 1981). Various measures of accessibility have been developed in the last decades. Arentze, et al.(1994) distinguish three groups of accessibility measures, namely, 'measures of travel costs', 'indicators of spatial opportunities' and 'measures of consumer welfare'. The first group includes all indicators measuring "the ease with which any facility can be reached from a specific location" (Arentze, et al., 1994), in terms of travel time, travel distance (can be further distinguished into network distance, and Euclidean distance, etc.), monetary charges and generalised cost (a combination of time and money costs). Indicators of spatial opportunities measure the opportunities available to particular demand locations, discounted by travel costs. Depending on the ways of coupling opportunities and travel costs, various measures, ranging from enumeration of opportunities at certain level of travel costs to weighted summation of opportunities using a negative travel cost function as a weight, can be derived. The whole family of Hansen-type measures (refered to as aggregate measures in Jones, 1981) and contour measures (refered to as disaggregate measures in Jones, 1981) belong to this group. Measures in the third group combine travel costs and attractiveness of supply into a subjective value (e.g. surplus, net benefit or utility). The underlying principle of these measures is the Marshall's measure of the change in consumer surplus accompanying a fall in the cost of a good. In this case, the goods to be considered are trips and the costs are travel costs. The measures in the first group concern only the transportation system, while the last two groups of measures combine in a single and simple measure the relevant characteristics both of land use and the transportation system.
An important drawback of all these measures, as pointed out by Arentze, et al, is their simplistic underlying assumption about people's travel behaviour, i.e. only single trip behaviour is assumed and trip chaining behaviour is ignored. Consequently, these measures are insensitive to certain spatial structures of supply, as illustrated in Arentze, et al, (1994). To overcome this shortcoming, they developed a so-called multistop measure of spatial opportunities, which is based on trip chaining behaviour. It is assumed that trip chaining behaviour occurs when a purchase fails at a destination, and continues until the intended purchase is realised. A so-called probability of success is used to determine the probability of terminating a trip chaining at a destination. Other important assumptions include that the probability of success at a next destination does not depend on the supply at former stops of the trip and the probability of success at a location which has been visited during the same trip is assumed to be zero. Based on these assumptions, a trip route is constructed along the way of minimum travel costs to the next destination until the predefined probability of success is met. The expected amount of travelling costs of the constructed trip route is then used as the indicator of accessibility(Areutze, et al., 1994). Although some of the assumptions of this approach may not be realistic, e.g., the assumption that the probability of success at a location which has been visited during the same trip is assumed to be zero. In reality, people intend to select the best location to purchase after visiting and comparing all potential locations available to them, which means the probability of success at a location which has been visited might be larger than zero However, the measure developed does account for the contribution of multi-stop trips to accessibility, and represents progress of research from simple trips to complex trips. Nevertheless, as admitted by the authors, the measure is confined to cases of multi-stop but single purpose (shopping) trips, therefore, it is not applicable to cases of multi-purpose and multi-stop trips.
Furthermore, this measure shares other shortcomings with the three groups of indicators discussed above For example, none of these measures accounts for the temporal constraints of individuals, nor for the opening hours of the facilities.
Consequently, these measures fail to capture the effects of opening hour changes and the implications of changes of individual's time schedules. While the spatial constraints are important factors to determine people's accessibility, the temporal constraints are equally if not more important elements to confine people's access to opportunities.
Association for European Transport