Social Sciences · Transportation
Exploring spatial non‐stationarity in the social determinants of time to work
See on the mapAbstract
Social differences in time to work are a growing concern of transport planning, as they can increase the risk of social exclusion of some groups. Recent work has found that those differences tend to vary across the urban space. Geographic Weighted Regression models this variability by estimating relationships in each place, weighting data by distance to that place. This paper builds on this work by assuming that households and policy-makers make decisions over space by comparing places according to their function within the metropolitan space, regardless of their geographic proximity. The determinants of time to work then vary in the space defined by variables identifying that function. This procedure is applied to the case of the Lisbon Metropolitan Area at two moments in time. Higher times to work are associated with less-qualified populations, with the strength of the relationships generally increasing with centrality in relation to employment locations and dependence on jobs in the main metropolitan centre. The priority to investments in road transport in the period concerned extended the disadvantages to second-ring suburban areas.