A longitudinal study of housing status and crime in a homeless population

Ann Clin Psychiatry. 2018 Nov;30(4):280-288.


Background: It is well established that homelessness is associated with crime. Previous research suggests that this association is at least partially secondary to homeless status offenses (eg, vagrancy and trespassing resulting from behaviors intrinsic to homelessness). To investigate this relationship, this study compared criminal behavior in a homeless population under housed and unhoused conditions.

Methods: Reported criminal and housing histories of 255 homeless individuals were examined at 3 annual interviews. Lifetime criminal charges incurred while homeless vs housed were compared for types of offenses. Prospective longitudinal data from these interviews were used to examine housing status and recent crime.

Results: At baseline, homeless status offenses were the most frequently reported charges and were the only type of crime for which charges were incurred more frequently in the homeless condition. Over the 2-year followup period, recent crime was consistently higher in groups who had been homeless relative to groups who had been housed, and crime rates fell after obtaining housing.

Conclusions: These findings suggest that homelessness itself is linked to criminal behavior through homeless status offenses. Negative effects of arrest and incarceration on housing acquisition warrant consideration of alternative legal system interventions to break the cycle of homelessness.

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Adult
  • Crime / statistics & numerical data*
  • Female
  • Housing / statistics & numerical data*
  • Humans
  • Ill-Housed Persons / statistics & numerical data*
  • Interviews as Topic
  • Longitudinal Studies
  • Male
  • Prospective Studies
  • Substance-Related Disorders / complications