Homelessness is an enduring problem, involving an estimated 550,000 Americans in 2016. Several intervention strategies have been employed since the 1990s to reduce its prevalence. Among the most widely-adopted of these was the 100,000 Homes Campaign, which took place between 2010 and 2014. While success has been claimed for each intervention, empirical tests of effectiveness are difficult to conduct, in part because data are hard to verify.

QUESTION. Did the 100,000 Homes Campaign reduce chronic homelessness?

METHODS. Structured interviews were conducted with experts in the homelessness field. A difference-in-differences regression was employed to test the effectiveness of the 100,000 Homes Campaign. Case studies were conducted of one state (Utah) and three communities (Orlando, FL; New Orleans, LA; and Santa Ana/Orange County, CA) which achieved large estimated reductions in chronic homelessness between 2009 and 2015.

RESULTS. Most interviewees believed the campaign was valuable, and most were optimistic that chronic homelessness could be ended within a few decades. Interviews revealed a striking lack of understanding of continuing inflow into chronic homelessness – that is, incidence rather than prevalence. The average chronically homeless rate for treatment group communities declined from 8.62 chronically homeless persons per 10,000 population in 2009 to 5.81 per 10,000 population in 2014, a decline of 32.6%. The average chronically homeless rate for control group communities declined from 5.07 per 10,000 in 2009 to 3.55 per 10,000 in 2014, a decline of 29.98 %. The difference between the two groups was not statistically significant (α > 0.05, -1.96 < t = 1.75 < +1.96). Case studies suggested potentially promising approaches, while demonstrating ongoing measurement challenges.

DISCUSSION. Contrary to expectations, 100,000 Homes Campaign had no statistically significant effect on reducing chronic homelessness. The availability of Veterans Assisted Supportive Housing (VASH) vouchers had the greatest effect on reduced prevalence. Advocacy for ending chronic homelessness and the analytics necessary to measure whether that outcome is being achieved are not well matched.