Trends in earnings and employment before and after the first instance of homelessness: A multi-cohort analysis

Although it has been suggested that homelessness has a persistently negative impact on earnings and employment, little empirical research exists to support this claim. We examined earnings and employment trends 10-years prior to and following the first instance of homelessness in order to better understand its economic impact. We explored how such trends were influenced by demographic and homelessness-related characteristics, such as homelessness typology (chronic/episodic/temporary), exit housing (permanent/non-permanent), and homelessness duration. Homelessness data were obtained from the NYC Department of Homeless Services while earnings and employment data were obtained from the Social Security Administration in aggregated form for 13 cohorts (1990-2002) comprising 67,400 sampling clusters (defined by cohort and follow-up period [1980-2007], demographic and homelessness characteristics). Piecewise mixed-effects models were used to examine trends in earnings and employment, with random effects for cohort. Statistical analyses were conducted separately for singles and families, weighted by cluster size, and adjusted for demographic characteristics. Results indicated a precipitous decline in both earnings and employment for both singles and families prior to homelessness. However, following homelessness, earnings recovered both for singles and families, while employment recovered only for families. In both cases, recovery was greater and more rapid for transitional and chronic homelessness (as compared to episodic) and for those exiting to permanent housing. Results suggest earnings and rate of employment are not universally static phenomena following homelessness, with recovery of both outcomes for most subgroups. However, increased intervention efforts should be focused on singles, particularly in terms of obtaining employment assistance.