Phantom of the Neighbourhood : Longitudinal Studies on Area-based Conditions and Individual Outcomes

Sammanfattning: This dissertation consists of three self-contained but interrelated empirical studies focusing on theoretical, empirical and political questions in the multidisciplinary field of neighbourhood effect research. Along with a comprehensive introductory essay, each study addresses questions concerning the potential influence of neighbourhood characteristics on individual social and economic outcomes at different life stages.Study I combines longitudinal register and survey data from the ‘golden era’ of Swedish welfare policy to evaluate a hypothesised impact of neighbourhood poverty during adolescence on a wide range of outcomes (including, but not limited to, educational and employment status) within a counterfactual model framework based on matching on propensity scores. Extensive empirical analyses indicate that, when two groups of children who are identical according to observed factors before age 10 (including household income, family structure and welfare receipt) live in different types of neighbourhood in adolescence, the outcome for those who grow up in a poor neighbourhood is not more likely to be worse than for those who grow up in a more affluent neighbourhood.Study II considers the maximum theoretical scope of unique neighbourhood influence experienced during the years of growth on individuals’ later life income and social assistance recipiency. A three-level hierarchical linear model is applied to simultaneously distinguish variation in the outcomes over time from variation that is attributable to differences between neighbourhoods. By utilising longitudinal register data derived from a birth cohort who grew up in Stockholm at a time when Swedish welfare policy ambitions were at a peak, this study attempts to estimate the long-term significance of neighbourhood origin in the Swedish setting. The analyses clearly show that prior place of residence accounts for an exceedingly modest proportion of the variation in cohort members’ subsequent income and receipt of social assistance.Study III explores the hypothesised negative impact of disadvantaged neighbourhood conditions, individual disadvantage, and degree of labour market establishment on levels of social trust. Using data from the Swedish Longitudinal Survey among Unemployed, ordered logit regression analyses indicate that low levels of social trust are contingent upon perceived neighbourhood disorder, personal powerlessness, perceived fear of victimisation, and accumulated episodes of temporary employment. The tentative results also indicate that neighbourhood disorder, powerlessness, and fear of victimisation interact, magnifying the negative impact on social trust.