Économie et Statistique n° 457-458 - Transport Survey
Do Households Make Trade-Offs between Cost of Housing and Cost of Transportation: A Response in the Greater Paris Region
In the Greater Paris Region, the high level of property prices results in a higher proportion of income being dedicated to housing than in the provinces. In this context, the rise in fuel prices could become a burden on household solvency. Some authors consider this risk to be accentuated in suburban and rural areas. Increasing distance from Paris is seen as inducing high transportation costs coming on top of already-high housing budgets. However, the monocentric model of the urban economy contradicts this assertion, if not totally at least to some extent. This model predicts that transportation and housing budgets are fungible, either partially or totally, according to the hypotheses that are chosen. High transportation costs in outlying areas are therefore thought to be offset by a lower housing budget. This study tests the predictions of the monocentric model concerning fungibility of housing and transportation budgets in the case of Greater Paris households. The information in the Global Transport Survey 2001-2002 can be used to devise a simple econometric test (significance of the transportation cost variable in the housing expenditure equation) that is new. For a given occupation status, income is the main determinant of the housing budget, with other household characteristics having an impact of secondary importance. The different transportation cost variables that were tested were non-significant or marginal in their influence. In particular, elasticity of the housing budget to the transportation budget was zero for first-time buyers and private-sector tenants, and slightly positive (0.025) in the social housing sector. These results reject the hypothesis of trade-offs between housing and transportation costs. Incidentally, the study found a positive link between housing expenditure and price per m². This brings into question the frequent choice of Cobb-Douglas utility functions to model household residential preferences.