ABSTRACT
This chapter investigates the effect of residential and job moving behaviour on commuting distance. Precise information about commuting distance or commuting time is missing in the data set, as only information on the municipality of residence and workplace of the worker is available. Market imperfections are thought to prevent households from adjusting the commuting distance to the optimal level. The very high correlation between the disturbances makes clear that pooling the data is less efficient and thus explains the higher standard errors for the estimates of the pooled data. Notice that pooling the data or using a panel model renders different estimates for the coefficients of the variables educational achievement, parents, and children. The policy implications are based on the notion that workers search for better jobs and residences. Policies, which aim at reducing commuting - example policies which reduce the costs of moving residence or moving job -, will therefore hardly be effective in the short run.
