Residential mobility and persistent neighborhood deprivation

Research output: Contribution to journalJournal articleResearchpeer-review

1 Citation (Scopus)


If policy-makers wish to intervene in neighborhoods to improve the socioeconomic status of the neighborhoods, it is important to know the residential dynamics of neighborhoods before starting any interventions. This paper describes the residential dynamics of persistent neighborhood deprivation over a period of 28 years. We examine small-scale, predefined, physically contiguous neighborhoods with public housing. We utilize unique longitudinal administrative data for the period 1986 to 2013 to follow individuals and their characteristics over time. With our unique data, we are the first to document that deprived neighborhoods function as a transit stage in residents’ housing trajectories. On average, residents’ unemployment rate increases before they enter a deprived neighborhood, and sharply declines two years before they move out. In line with the aforementioned findings, the results suggest that an increase in economic resources is positively associated with leaving a deprived neighborhood compared with non-deprived neighborhoods. Hence, residents’ moving patterns contribute to deprived neighborhoods maintaining their relative position. However, as both the socioeconomic difference between in-movers and out-movers and the turnover rate decline over time, residents’ moving patterns account for a small share of persistent neighborhood deprivation. Thus, with the unique data, we show that deprivation in neighborhoods is mainly attributable to long-term residents who on average do not improve their socio-economic characteristics.
Original languageEnglish
Article number101771
JournalJournal of Housing Economics
Publication statusPublished - Sep 2021


  • Neighborhood deprivation
  • Moving patterns
  • Transit
  • Labor market attachment


Dive into the research topics of 'Residential mobility and persistent neighborhood deprivation'. Together they form a unique fingerprint.

Cite this