Table 3.
Cox Proportional Hazard Regression Model Results for Shelter Exit1.
| Variables | Model 1 | Model 2 |
|---|---|---|
| Individual factor variables | ||
| Male | 1.064 (0.073) | 0.775 * (0.075) |
| Race | ||
| White | 1.075 (0.103) | 0.853 (0.105) |
| Black | 0.949 (0.115) | 0.843 (0.118) |
| Indian/Native American | 1.004 (0.120) | 0.881 (0.124) |
| Hispanic 2 | 0.969 (0.051) | 0.970 (0.052) |
| Age in shelter entry | 0.998 (0.003) | 1.002 (0.003) |
| Household size | 0.904 *** (0.020) | 0.973 (0.021) |
| Adults | 0.790 *** (0.047) | 0.772 *** (0.050) |
| Children less than 5 years old | 1.051 (0.031) | 1.044 (0.032) |
| Veteran 2 | 1.217 (0.157) | 1.504 * (0.159) |
| Disability 2 | 1.039 (0.054) | 0.993 (0.055) |
| Alcohol abuse 2 | 1.343 (0.144) | 1.637 ** (0.148) |
| Chronic health 2 | 0.977 (0.058) | 0.912 (0.059) |
| Drug abuse 2 | 0.736 ** (0.095) | 0.714 * (0.100) |
| Mental health 2 | 0.821 *** (0.057) | 0.912 (0.058) |
| Domestic violence occurred 3 | ||
| Unknown | 1.079 (0.065) | 1.100 (0.067) |
| Less than 3 months ago | 0.884 (0.074) | 0.975 (0.075) |
| 3–6 month ago | 0.655 ** (0.134) | 0.789 * (0.069) |
| 6 month–1 year ago | 0.675 *** (0.122) | 0.711 ** (0.136) |
| More than 1 year ago | 0.661 *** (0.068) | 0.683 *** (0.069) |
| Structural factor variables | ||
| Housing program enrollment during the homeless episode | ||
| Emergency shelter | 0.463 *** (0.098) | |
| Rapid rehousing | 0.210 *** (0.060) | |
| Permanent supportive housing | 0.016 *** (0.370) | |
| Transitional housing | 0.520 *** (0.181) | |
| Prior residence | ||
| Emergency shelter/hotel, motel vouchers 4 | ||
| Less than 1 month | 1.087 (0.074) | |
| 1 month–90 days | 0.999 (0.105) | |
| 90 days–1 year | 0.942 (0.147) | |
| More than 1 year | 1.546 (0.260) | |
| Housing by a homeless family 4 | ||
| Less than 1 month | 1.611 (0.508) | |
| 1 month–90 days | 1.618 *** (0.232) | |
| 90 days–1 year | 1.115 (0.135) | |
| More than 1 year | 1.573 ** (0.140) | |
| Double up 4 | ||
| Less than 1 month | 1.264 ** (0.066) | |
| 1 month–90 days | 1.044 (0.096) | |
| 90 days–1 year | 1.128 (0.125) | |
| More than 1 year | 1.663 ** (0.132) | |
| Average monthly income ($100) | 1.008 (0.007) | |
| Average earned income ($100) | 0.981 * (0.008) | |
| Average family support income ($100) | 0.968 ** (0.010) | |
| Average rental assistance ($100) | 0.855 *** (0.011) | |
| Season when entering the episode 5 | ||
| Summer (June to August) | 1.041 (0.062) | |
| Fall (September to November) | 1.045 (0.066) | |
| Winter (December to February) | 1.024 (0.066) | |
| Coming from Salt Lake County 6 | ||
| No | 1.356 *** (0.074) | |
| Yes | 1.192 *** (0.057) | |
| n | 2348 | 2348 |
| R2 | 0.082 | 0.531 |
| Log Likelihood | −14,409.76 | −13,622.00 |
* p < 0.05; ** p < 0.01; *** p < 0.001; 1 The dependent variable here is “shelter exit”—a binary variable representing whether a homeless family stayed the shelter or not. 2 The reference group is a “no.” 3 The reference group is “No domestic violence occurred.” 4 The reference group is “not staying before entering the shelter.” 5 “Spring” is a reference group for the season of the shelter entry variable. 6 “Unknown” is a reference group for the variable.