Residential location, workplace location and car driving in four Norwegian cities

ABSTRACT

Based on a study in four Norwegian cities (Oslo, Stavanger/Sandnes, Bergen and Trondheim) differing in size and centre structure, this article illuminates how residential and workplace location, local-area density and transit accessibility influence different aspects of travel behaviour. We find strong effects of residential and workplace distance to the city centre on overall driving distances and commuting distances. We also find clear effects of local area densities around residences and workplaces on the choice of car as a travel mode, along with less pronounced effects of the distance from dwellings and workplaces to the city centre. In the cities with the best developed transit provision, we also see clear effects of transit accessibility at the residence on the propensity of choosing the car as travel mode. The results provide strong support of Norwegian national policies of urban densification as a planning strategy to curb the growth in urban motoring. However, although the influences of urban structure on travel show many similarities across the four cities, there are also important differences reflecting variations in centre structure (predominantly mono- or polycentric) and population size. The magnitude of the influences of various urban structural characteristics on travel behaviour are thus highly context-dependent.

KEYWORDS:

• Residential location
• workplace location
• car driving
• monocentric
• polycentric
• four cities

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 This indicator might not in itself distinguish unambiguously between a polysentric and a sprawled urban structure. However, the large difference between Stavanger and the three other cities in employment distribution cannot be attributed merely to different degrees of sprawl. As mentioned above, there is great similarity in population densities between the three medium-sized cities (Stavanger, Bergen and Trondheim), ranging from 29 to 31 persons per hectare. Instead of representing sprawl, the low monocentric bias value for jobs in Stavanger thus reflects the strongly polycentric distribution of jobs over the urban area.

2 In Norway, the location of new development is decided first in the municipal master land use plan differentiating between developmental areas and areas where no construction is to take place. The density and design of the planned development is decided in subsequent, more detailed local development plans before final building permits are given based on architectural drawings. The local development plans must be in accordance with the municipal master land use plans, and although exemptions are sometimes given, there is normally conformance between the two tiers of planning.

3 Distances to the various categories of centres are measured as the fastest driving route along the road network. This also applies to commuting distances as dependent variables.

4 In the city region of Trondheim, the closest second-order centres comparable to those in Oslo, Bergen and Stavanger (Stjørdalshalsen and Orkanger) are located 34 km and 42 km, respectively, away from the city centre of Trondheim.

5 Measured within a 750 × 750 m square centred on the 250 × 250 m square to which the address belongs.

6 Departures per hour within 1.5 km from the dwelling between 07 and 09 on weekdays.

7 The study by Cervero and Duncan (2006) investigated jobs-housing balance at a relatively large geographical scale (within 4 km from the dwelling) but did not control for distance to metropolitan city centre.

Additional information

Funding

This work was supported by Norges Forskningsråd: [Grant Number 23013].