Commuting as a rural growth strategy? : evidence from Sweden

The aim of this paper is to examine the extent to which rural regions benefit from closeness to large cities. It is often argued that rural regions can enjoy urban-based growth through urban sprawl, i.e. deconcentration of urban activities to rural regions. Analyses of rural regions in e.g. Canada show that thriving and prosperous regions (in terms of employment and population growth) often are located within commuting distance to urban centers (e.g. Partridge et al 2010). This suggests that one mechanism behind these patterns involves the process of people moving to the countryside while working in the city (Betrand & George-Marcelpoil 2005), which put rural-to-urban commuters at the center of attention (cf. Goetz et al 2010). Even if this type of commuting flows are prime examples of rural-urban integration and a main way in which rural regions may enjoy urban spread effects, there are few analyses that directly examines rural-to-urban commuting (Partridge et al 2010). Many key issues are not fully analyzed. For instance; what are the critical distances for urban spread effects through rural-to-urban commuting? To what extent do they differ across different categories of the labor force and across different sectors? In this paper, these questions are examined using detailed Swedish longitudinal matched employer-employee data. The main results show that regions close to urban centers experience positive spread effects from closeness to large cities. Evidence shows that rural population growth from urban sprawl is a key determinant behind rural-to-urban commuting. Using the methodology developed by Partridge et al (2010), the paper also estimate how distance to nearest urban center influence commuting patterns; thereby capturing the rural-urban interdependence. A novelty in the analysis is that critical distances where urban spread effects for rural regions vanishes are estimated for different parts of the labor force with respect to education level and occupation. The main results show that there exist small differences in distance between these groups.