People in Supportive Housing Used Less Emergency Health Care and Received More Office-Based Care on Average than People Receiving Usual Services

Urban Institute

Sources: Colorado Access, Denver Health and Hospital Authority, Denver Police Department, and Denver Sheriff Department.

Notes: SIB=Denver Supportive Housing Social Impact Bond. Sample for the treatment group is 275 people. Sample for the control group is 274 people. Results are for two years after people were randomized to the treatment and control groups. Results were estimated using ordinary least squares. The regression-adjusted models included the following control measures: age, gender, race/ethnicity, days in jail in the three years prior to randomization, and number of arrests in the three years prior to randomization. In addition, each regression controlled for the value of the outcome in the year before randomization. The differences are statistically significant at the 0.01 level.