Contextual poverty and obtained educational level and income in Sweden and the Netherlands: A multi-scale and longitudinal study

Abstract

Studies of neighbourhood effects typically measure the neighbourhood context at one specific spatial scale. It is increasingly acknowledged, however, that the mechanisms through which the residential context affects individual outcomes may operate at different spatial scales, ranging from the very immediate environment to the metropolitan region. We take a multi-scale approach to investigate the extent to which concentrated poverty in adolescence is related to obtained education level and income later in life, by measuring the residential context as bespoke neighbourhoods at five geographical scales that range from areas encompassing the 200 nearest neighbours to areas that include the 200k+ nearest neighbours. We use individual-level geocoded longitudinal register data from Sweden and the Netherlands to follow 15/16-year-olds until they are 30 years old. The findings show that the contextual effects on education are very similar in both countries. Living in a poor area as a teenager is related to a lower obtained educational level when people are in their late 20s. This relationship, however, is stronger for lower spatial scales. We also find effects of contextual poverty on income in both countries. Overall, this effect is stronger in the Netherlands than in Sweden. Partly, this is related to differences in spatial structure. If only individuals in densely populated areas in Sweden are considered, effects on income are similar across the two countries and income effects are more stable across spatial scales. Overall, we find important evidence that the scalar properties of neighbourhood effects differ across life-course outcomes.